The Mismatch Between Phenotypes and Genotypes in Lichenized Fungi (Bryoria Sect

Persoonia 42, 2019: 75–100 ISSN (Online) 1878-9080 www.ingentaconnect.com/content/nhn/pimj RESEARCH ARTICLE https://doi.org/10.3767/persoonia.2019.42.04

Evaluating methodologies for species delimitation: the mismatch between phenotypes and genotypes in lichenized fungi (Bryoria sect. Implexae, Parmeliaceae)

C.G. Boluda1,2, V.J. Rico1, P.K. Divakar1, O. Nadyeina2, L. Myllys3, R.T. McMullin4, J.C. Zamora1,5, C. Scheidegger 2, D.L. Hawksworth6

Key words Abstract In many lichen-forming fungi, molecular phylogenetic analyses lead to the discovery of cryptic species within traditional morphospecies. However, in some cases, molecular sequence data also questions the separation chemotypes of phenotypically characterised species. Here we apply an integrative taxonomy approach ‒ including morphologi- cryptic species cal, chemical, molecular, and distributional characters ‒ to re-assess species boundaries in a traditionally speciose haplotypes group of hair lichens, Bryoria sect. Implexae. We sampled multilocus sequence and microsatellite data from 142 incomplete lineage sorting specimens from a broad intercontinental distribution. Molecular data included DNA sequences of the standard fungal integrative taxonomy markers ITS, IGS, GAPDH, two newly tested loci (FRBi15 and FRBi16), and SSR frequencies from 18 microsatellite microsatellites markers. Datasets were analysed with Bayesian and maximum likelihood phylogenetic reconstruction, phenogram speciation reconstruction, STRUCTURE Bayesian clustering, principal coordinate analysis, haplotype network, and several dif- species concepts ferent species delimitation analyses (ABGD, PTP, GMYC, and DISSECT). Additionally, past population demography and divergence times are estimated. The different approaches to species recognition do not support the monophyly of the 11 currently accepted morphospecies, and rather suggest the reduction of these to four phylogenetic species. Moreover, three of these are relatively recent in origin and cryptic, including phenotypically and chemically variable specimens. Issues regarding the integration of an evolutionary perspective into taxonomic conclusions in species complexes, which have undergone recent diversification, are discussed. The four accepted species, all epitypified by sequenced material, are Bryoria fuscescens, B. glabra, B. kockiana, and B. pseudofuscescens. Ten species rank names are reduced to synonymy. In the absence of molecular data, they can be recorded as the B. fuscescens complex. Intraspecific phenotype plasticity and factors affecting the speciation of different morphospecies in this group of Bryoria are outlined.

Article info Received: 21 December 2017; Accepted: 18 June 2018; Published: 23 August 2018.

INTRODUCTION of delimited clades (Jakob & Blattner 2006, Lumley & Sperling 2011), or incomplete lineage sorting (Saag et al. 2014, Leavitt Accurate identification and characterization of species is the et al. 2016). Long-term reproductive isolation may produce basis of communication, conservation, resources management, structured, non-overlapping lineages, whereas an intraspecific and material used in biological research. However, in groups phylogeny, as well as a recent or contemporary speciation of relatively recent origin, species delimitation is often difficult event, may produce reticulated lineages (Abbott et al. 2016). (Jakob & Blattner 2006, Leavitt et al. 2011, Lumley & Sperling The family Parmeliaceae is one of the most studied amongst 2011). Organisms are always evolving, changing in response to lichenised fungi. It contains many genera with species delimita- either selective pressures or genetic drift, so that delimiting units tion problems, such as Cetraria aculeata (Lutsak et al. 2017); to accord species names is not always clear (Naciri & Linder Letharia (Altermann et al. 2014), the Parmotrema reticulatum 2015). Several phenomena can hinder species delimitation: complex (Del-Prado et al. 2016), and Pseudephebe (Boluda et phylogenetic/phenotypic mismatches (Articus et al. 2002, Mark al. 2016). In some cases, a lack of correlation between geno- et al. 2016, Pino-Bodas et al. 2016), ‘intermediate’ specimens types and phenotypes has led to the recognition of cryptic between generally accepted taxa (Seymour et al. 2007), hybridi- species within morphologically indistinguishable or scarcely zation (Konrad et al. 2002, Steinová et al. 2013), an absence indistinguishable morphospecies (Molina et al. 2011a, b, Leavitt 1 Departamento de Farmacología, Farmacognosia y Botánica (U.D. Botá et al. 2012a, b, Singh et al. 2015, Boluda et al. 2016, Del-Prado nica), Facultad de Farmacia, Universidad Complutense, Plaza de Ramón et al. 2016), and so far, more than 80 cryptic lineages have been y Cajal s/n, Madrid 28040, Spain; detected in Parmeliaceae (Crespo & Lumbsch 2010, Divakar et corresponding author e-mail: Carlos G. Boluda, [email protected]. 2 Biodiversity and Conservation Biology, Swiss Federal Research Institute al. 2010). However, in other cases there is a mismatch between WSL, Zürcherstrasse 111, Birmensdorf 8903, Switzerland. lineages revealed by standard DNA-barcoding markers and 3 Botanical Museum, Finnish Museum of Natural History, P.O. Box 7, 00014 long-accepted morphospecies (Articus et al. 2002, Seymour University of Helsinki, Finland. et al. 2007, Velmala et al. 2014, Mark et al. 2016, Kirika et al. 4 Research and Collections, Canadian Museum of Nature, Ottawa, ON K1P 6P4, Canada. 2016a, b, McMullin et al. 2016). 5 Museum of Evolution, Uppsala University, Norbyvägen 16, 75236 Uppsala, In the morphologically similar ‘beard’ and ‘hair’ lichens of the Sweden. Alectoria sarmentosa, Bryoria sect. Implexae, and Usnea 6 Department of Life Sciences, The Natural History Museum, Cromwell Road, London SW7 5BD, UK; and Comparative Plant and Fungal Biology, Royal barbata species complexes (Velmala et al. 2014, Mark et al. Botanic Gardens, Kew, Surrey TW9 3DS, United Kingdom. 2016, McMullin et al. 2016), DNA sequences from standard

© 2018-2019 Naturalis Biodiversity Center & Westerdijk Fungal Biodiversity Institute You are free to share - to copy, distribute and transmit the work, under the following conditions: Attribution: You must attribute the work in the manner speciﬁed by the author or licensor (but not in any way that suggests that they endorse you or your use of the work). Non-commercial: You may not use this work for commercial purposes. No derivative works: You may not alter, transform, or build upon this work. For any reuse or distribution, you must make clear to others the license terms of this work, which can be found at http://creativecommons.org/licenses/by-nc-nd/3.0/legalcode. Any of the above conditions can be waived if you get permission from the copyright holder. Nothing in this license impairs or restricts the author’s moral rights. 76 Persoonia – Volume 42, 2019 ‒ KY002700 KY002731 ‒ ‒ KY002730 KY002688 KY002692 ‒ KY002702 ‒ ‒ ‒ ‒ ‒ ‒ ‒ ‒ ‒ KY002732 KY002699 KY002701 KY002742 ‒ ‒ KY002693 KY002691 ‒ KY002690 ‒ KY002698 ‒ ‒ ‒ KY002716 KY002694 KY002695 ‒ ‒ KY002720 KY002718 KY002697 KY002711 KY002715 KY002717 KY002696 KY002719 KY002751 KY026820 KY026821 KY026813 KY026814 ‒ KY026825 KY026817 KY026819 KY026822 KY026823 KY026818 ‒ ‒ ‒ ‒ ‒ KY026816 KY026815 KY026824 KY026802 KY026800 KY026803 ‒ KY026810 KY026807 KY026806 KY026809 KY002729 KY026811 KY026801 KY026805 KY026804 KY026808 KY026812 ‒ KY026994 KY026995 KY026955 KY026996 KY026997 KY026998 KY026999 KY027000 KY027001 KY026988 KY026986 KY026989 ‒ KY026992 KY026987 KY026991 KY026990 KY026993 GAPDH FRBi15 FRBi16 GenBank accession numbers KY026947 KY026948 KY083556 KY026949 KY026950 KY026951 KY026952 KY026953 KY026954 KY026941 KY026939 KY026942 KY083555 KY026945 KY026940 KY026944 KY026943 KY026946 IGS KJ396437 KJ396499 KJ954310 KJ396438 ‒ KY026902 KY026903 KY026909 KJ396441 GQ996307 KJ396506 KJ396505 KJ954313 GQ996280 ‒ KY026901 GQ996305 KJ396498 GQ996278 KJ396440 GQ996304 KJ396503 KJ396500 KJ954312 GQ996277 ‒ GQ996306 GQ996300 KJ396504 KJ396497 KY026904 GQ996279 GQ996272 KY026905 KY026906 KY026907 KY026908 FJ668494 FJ668456 FJ668400 ‒ HQ402722 KJ396494 HQ402627 KJ396442 ‒ KJ396443 KJ396507 KJ576716 KJ396509 KJ576717 KJ954314 KJ396510 KJ396444 KJ954315 KJ396511 ‒ KJ396445 KJ599469 KJ396514 ‒ KJ396446 KJ599470 KJ396515 ‒ GQ996303 KJ954316 KJ396516 ‒ KJ954317 KJ396512 ‒ KJ954318 ‒ GQ996276 ‒ ‒ GQ996290 KJ396436 KJ396496 KJ576715 KJ396495 GQ996262 GQ996291 KJ396501 KJ954309 KJ396502 KJ599468 GQ996263 ‒ KY026895 KY026893 KY026896 ‒ KJ576728 KJ396493 KJ599481 ‒ GQ996289 KJ396490 GQ996261 KY026899 FJ668493 FJ668455 FJ668399 GQ996287 KJ396433 KJ396487 KY026894 KJ396489 GQ996259 KJ954306 GQ996288 KJ396488 GQ996260 ‒ KY026898 KY026897 KY026900 KJ396435 KJ396492 KJ954308 KJ396434 KJ396491 KJ954307 ‒ ITS

1 Fum. Fum. Fum. Fum. Abs. Fum. Fum. Fum. Fum. Fum. Abs. Fum. Fum. Fum. Fum. Fum. Fum. Fum. Abs. Fum. Fum. Fum. Fum. Fum. Fum. Fum. Fum. Abs. Fum. Fum. Fum. Fum. Ale., Bar. Ale., Bar. Ale., Bar. Gyr. Gyr. Ale., Bar. Ale., Bar. Ale., Bar. Ale., Bar. Ale., Bar Ale., Bar. Bar. Fum., Pso. Ale., Bar., Ale., Bar. Pso. Nor., Ale., Bar., Gyr. Gyr.

2 5 6 7 8 9 1 ‒ 64 66 81 82 71 83 70 63 67 65 62 68 72 73 74 75 77 78 79 76 60 69 80 61 90 89 92 17 16 87 95 84 85 88 91 86 94 93 96 15 14 Appendix 4 code Fig. 2 and Chemistry L224 L305 L12.03 L12.05 S259 L15.21 S256 L189 S109 L232 L160 Bg1 S157 Bg2 Bg3 Bg4 Bg5 L186 L147 S260a S261 S267 S272 S369 S379 S380 S274 L139 S24 S56 L149 L07.15 L01.17 L08.12 fri_02 S395a Lab. code S192 L15.15 L141 L211 S2 L06.10 L270 L14.02 L13.03 L16.21 L407 L355 et al. (2014) Velmala et al. (2014) Velmala MAF-Lich. 19681 MAF-Lich. 19682 (UBC) Dillman 11May11:1 L406 et al. (2014) Velmala MAF-Lich. 20684 et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala MAF-Lich. 20595 et al. (2014) Velmala MAF-Lich. 20596 MAF-Lich. 20597 MAF-Lich. 20598 MAF-Lich. 20599 et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala MAF-Lich. 19672 MAF-Lich. 19664 MAF-Lich. 19674 MAF-Lich. 20602 et al. (2014) Velmala

et al. (2014) Velmala Source/Voucher MAF-Lich. 20683 et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala MAF-Lich. 19670 et al. (2014) Velmala MAF-Lich. 20682 MAF-Lich. 19685 MAF-Lich. 20687 et al. (2014) Velmala et al. (2014) Velmala Tenerife Tenerife Tenerife Territory Alberta Alberta Alberta Alberta Alberta Alberta Alberta Alberta Alberta Telemark Troms Troms Ahvenanmaa Teruel Alaska (epitype) Alaska Canada, Sogn og Fjordane Norway, Norway, Norway, Norway, Russia, Perm USA, Canada Canada, Canada, Canada, Canada, Canada, Canada, Canada, Canada. Canary Islands, Finland, Finland, Oulun Pohjanmaa Finland, Pohjois-Karjala Greenland Sweden, Södermanland Chile, IX Region Chile, IX Region Chile, IX Region Chile, IX Region Chile, IX Region Finland, Koillismaa Finland, Etelä-Savo Finland, Etelä-Savo (epitype) Finland, Koillismaa Finland, Koillismaa et al. (2014) Velmala Spain, Lérida Spain, Madrid Spain, Navarra Spain, Canada, British Columbia USA, Locality Canary Islands, Canary Islands, Finland, Etelä-Häme Finland, Etelä-Savo Finland, Uusimaa Greece, Peloponese Nord-Trøndelag Norway, Sweden, Västerbotten Switzerland, Bivio Canada Canada, British Columbia

B. fuscescens

B. glabra

B. furcellata

Newly obtained sequences are in bold . 1 Specimen information and GenBank accession numbers of the Bryoria sect. Implexae samples used in this study. Table Taxon Bryoria capillaris

B. friabilis

C.G. Boluda et al.: Mismatch between phenotypes and genotypes in lichenized fungi 77 KY002689 ‒ KY002704 KY002714 ‒ KY002733 ‒ ‒ ‒ KY002764 KY002765 KY002721 KY002705 ‒ KY002761 ‒ KY002756 ‒ ‒ KY002749 KY002745 KY002727 KY002724 ‒ KY002703 KY002760 ‒ ‒ ‒ ‒ KY002743 KY002744 ‒ ‒ ‒ ‒ KY002762 KY002753 ‒ KY002754 KY002755 ‒ ‒ KY002706 ‒ ‒ KY002752 KY002759 ‒ ‒ KY026829 KY026832 KY026833 KY026834 KY026835 KY026827 KY026840 KY026841 KY026828 KY026831 KY026837 KY026863 ‒ ‒ KY026857 KY026858 KY026859 KY026861 ‒ KY026826 KY026836 KY026838 KY026839 KY026846 KY026843 KY026844 KY026830 KY026847 KY026850 KY026851 KY026852 KY026853 KY026849 KY026842 KY026848 KY026845 KY026864 ‒ KY026956 GAPDH FRBi15 FRBi16 KY027005 KY027003 KY027004 ‒ ‒ KY027014 KY027015 KY027016 ‒ ‒ KY027002 KY027008 KY027009 KY027006 KY027007 KY027010 GenBank accession numbers ‒ KY083557 IGS KY026960 KY026958 KY026959 ‒ ‒ KY026971 KY026972 KY026973 KY026975 KY026965 KY026957 KY026963 KY026964 KY026961 KY026962 ‒ KY026911 KY026910 ITS KY026915 GQ996294 KJ576719 KJ396517 KJ396518 GQ996266 KJ599472 GQ996293 KJ396519 GQ996265 KJ396447 KJ396520 KJ954319 KY026913 KJ396453 KJ396454 KJ396529 KJ396530 KJ954325 KJ954326 KY026914 KJ396448 KJ396521 KJ954320 KJ396450 KJ396524 KJ954322 KJ576724 KJ396527 KJ599479 ‒ KJ396465 ‒ KJ396546 KJ954337 ‒ KY026925 KY026926 KY026927 ‒ KY026921 KY026912 KJ396449 KJ396523 KJ954321 KJ396451 GQ996284 KJ396525 KJ396452 KJ396526 KJ954323 KJ396528 GQ996256 KJ954324 ‒ GQ996295 KY026918 KJ396531 KY026919 GQ996267 KJ396464 KJ396545 KJ954336 ‒ KY026916 GQ996283 KJ396522 GQ996255 ‒ GQ996296 KJ396532 GQ996268 KJ396462 KJ396459 KJ396543 KJ396460 KJ396540 KJ396461 KJ954334 KJ396541 KJ396463 KJ954331 KJ396542 KJ954332 KJ396544 KJ954333 KJ954335 ‒ KJ396455 KJ396456 KJ396533 KY026917 KJ396535 KJ576720 KJ954327 KY026920 KJ954328 KJ396534 ‒ KJ599473 KJ576714 KJ396539 KJ599467 ‒ KJ396466 KJ396547 KJ954338 KJ396457 KJ396536 KJ954329 ‒ KJ396467 KJ396548 KJ954339 ‒

1 Fum. Abs. Pso. Pso. Pso. Pso. Pso. Fum., Pso. Pso. Pso. Pso. Pso. Abs. Abs. Ale., Bar. Ale. Ale., Bar. Ale., Bar. Ale., Bar. Ale., Bar. Ale., Bar. Ale., Bar. Pso. Abs. Abs. Abs. Abs. Nor. Nor. Nor. Ale., Bar. Pso. Abs. Nor. Ale. Ale., Gyr. Ale., Gyr. Ale., Gyr. Ale., Bar. Nor. Nor. Nor. Nor. Pso. Nor., Ale., Bar. Ale., Bar. Ale., Bar. Ale., Bar.

4 3 11 98 99 10 97 20 23 33 51 49 38 39 40 42 43 19 21 22 24 32 18 27 28 29 30 31 26 34 25 35 111 113 114 112 110 115 105 100 101 103 104 102 107 106 108 109 Appendix 4 code Fig. 2 and Chemistry S388 Lab. code L11.15 S22 S36 S39 S67 L06.05 L10.03 L394 L396 S168 L01.01 L347 S384 L244a S368 pik_c pik_a pik_02 pik_04 pik_05 pik_09 pik_10 L16.15 L323b L358 S239a S392a L09.04 L09.07 L274 S362 L206 L275 L421 L374 L376 L377 S221 S160 L04.03 S128 L16.17 L210 S382 L197 S383a (UBC)

Source/Voucher (UBC) Dillman 26July11:4 L414 Björk 1546 MAF-Lich. 19683 et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala MAF-Lich. 19669 MAF-Lich. 19679 et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala MAF-Lich. 19663 et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala MAF-Lich. 20601 MAF-Lich. 20600 MAF-Lich. 20685 et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala MAF-Lich. 19677 MAF-Lich. 19678 et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala MAF-Lich. 19667 et al. (2014) Velmala MAF-Lich. 20686 et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala Atlas Atlas Territory Azarbaijan Alberta Alberta Troodos Alaska Alaska (holotype) Alaska Alaska Cyprus, Locality USA, USA, Washington Finland, Koillismaa Finland, Koillismaa Finland, Koillismaa Finland, Koillismaa Greece, Peloponese Morocco, Rif USA, USA, Iran, East Russia, Murmansk Spain, Madrid Canada, British Columbia (holotype) et al. (2014) Velmala USA, Canada, British Columbia Canada, British Columbia Canada, Nova Scotia Canada, Prince Edward Island MAF-Lich. 20603 Canada, Prince Edward Island MAF-Lich. 20606 Canada, Prince Edward Island MAF-Lich. 20607 Canada, Prince Edward Island MAF-Lich. 20609 Canada, Prince Edward Island MAF-Lich. 20612 Switzerland, Bivio Canada, British Columbia Canada, British Columbia Canada, British Columbia Canada, British Columbia Morocco, Middle Morocco, Middle Nord-Trøndelag Norway, Canada, British Columbia Canada, British Columbia Nord-Trøndelag Norway, Russia, Perm Canada, British Columbia Canada, British Columbia Canada, British Columbia Canada, British Columbia Canada, British Columbia Spain, Zamora Sweden, Härjedalen Switzerland, Bivio Canada, Canada, British Columbia Canada, Canada, British Columbia B. implexa Table 1 (cont.) Table Taxon B. glabra (cont.)

B. kockiana B. kuemmerleana

B. inactiva

B. pikei

78 Persoonia – Volume 42, 2019 KY002707 KY002758 KY002766 KY002763 ‒ ‒ KY002739 ‒ ‒ ‒ KY002757 ‒ KY002740 ‒ ‒ KY002710 KY002750 ‒ KY002726 KY002728 KY002725 KY002723 KY002741 KY002748 ‒ ‒ ‒ ‒ ‒ ‒ KY002709 ‒ KY002746 KY002722 KY002747 ‒ KY002738 KY002712 KY002713 KY002708 KY002734 KY002735 KY002736 KY002737 ‒ ‒ KY026867 KY026869 KY026868 KY026886 KY026885 KY026865 KY026866 KY026892 KY026882 KY026887 KY026854 ‒ KY026855 KY026856 ‒ KY026860 ‒ KY026862 KY026888 KY026889 KY026884 KY026890 KY026891 KY026872 KY026881 KY026877 KY026880 KY026878 KY026883 KY026871 KY026870 KY026873 KY026874 KY026875 KY026876 KY026879 KY027011 ‒ KY027012 KY027013 ‒ KY027017 ‒ KY027018 GAPDH FRBi15 FRBi16 KY027021 KY027026 KJ954311 ‒ KY027027 KY027020 KY027019 KY027022 KY027023 KY027024 KY027025 KY027028 GenBank accession numbers KY026966 KY026967 KY026968 KY026969 ‒ KY026974 KY026970 ‒ IGS KY026978 KY026983 KY026984 KY026977 KY026976 KY026979 KY026980 KY026981 KY026982 KY026985 KJ576726 KJ396557 KJ396486 KJ396485 KJ599477 KJ396581 KJ396580 KJ954358 KJ954357 KJ576725 KJ396556 KJ599478 ‒ KJ396473 KJ396555 KJ954345 ‒ KJ396484 KJ396578 KJ954356 KJ396472 KJ396554 KJ954344 ‒ KJ396483 KJ396577 KJ954355 KJ396469 KJ396551 KJ954341 KJ396470 KJ396471 KJ396552 KJ396553 KJ954342 KJ954343 ‒ KJ396481 GQ996297 KJ396573 KJ396564 KJ954353 GQ996269 KJ576727 KJ396550 KJ599480 ‒ KJ396478 KJ396566 KJ954350 KY026922 ‒ KY026923 KY026924 ‒ ‒ ‒ KY026928 KJ396468 KJ396549 KJ954340 ‒ GQ996302 KJ396568 GQ996275 ITS KJ576722 KJ396480 KJ396570 KJ396571 KJ599475 KJ954352 ‒ KJ396482 KJ396576 KJ954354 KJ396477 KJ396563 KJ954349 KJ576721 KJ396572 KJ599474 KY026931 GQ996301 KJ396562 GQ996274 KY026936 GQ996299 KJ396474 KJ396558 KJ396559 GQ996271 KJ396439 KJ954346 ‒ ‒ KY026937 GQ996285 KJ396574 GQ996257 GQ996308 KJ396575 GQ996273 ‒ KY026930 KY026929 KY026932 KY026933 KY026934 KY026935 KY026938

1 Nor. Abs. Abs. Nor. Nor. Fum. Nor. Abs. Nor. Nor. Nor. Abs. Gyr. Ale., Bar. Gyr. Ale., Bar. Ale., Bar. Ale., Bar. Ale., Bar. Ale., Bar. Ale., Bar. Ale., Bar. Ale., Bar. Ale., Bar. Abs. Fum. Fum. Abs. Fum. Gyr. Fum. Abs. Abs. Gyr. Abs. Gyr. Fum., Gyr. Abs. Fum. Fum. Fum. Fum. Abs. Abs. Fum. Gyr.

59 12 13 58 57 56 53 54 55 37 44 45 46 47 52 41 48 50 36 119 118 116 117 124 123 131 125 126 127 128 122 130 129 135 140 132 141 120 121 134 133 136 137 138 139 142 Appendix 4 code Fig. 2 and Chemistry S387 L395 L392 S386 S385 S377 S341b S371 S72 S222 S232 S370 S10 S42 S394 S45 Lab. code pik_11 pik_12 pik_13 pik_14 pik_d pik_07 pik_15 pik_b S390 S57 S59 S196a S6 S62 L05.17 L10.13 L300 L307 L272 L273 L12.11 S164 S166 L03.07 L02.20 L07.03 L07.19 L08.19 L08.20 L13.12 et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala

Source/Voucher MAF-Lich. 20608 MAF-Lich. 20605 MAF-Lich. 20604 et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala MAF-Lich. 19668 MAF-Lich. 19680 et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala et al. (2014) Velmala MAF-Lich. 19684 et al. (2014) Velmala et al. (2014) Velmala MAF-Lich. 19666 MAF-Lich. 19665 MAF-Lich. 19671 MAF-Lich. 19673 MAF-Lich. 19675 MAF-Lich. 19676 MAF-Lich. 19686 Territory Territory = Norstictic acid; Pso. Psoromic acid. = Gyrophoric acid; Nor. = Barbatolic acid; Fum. Fumarprotocetraric Gyr. Alectorialic acid; Bar. Alberta Troms Asturias Alaska Alaska Alaska Alaska Alaska Alaska Ale. = USA, USA, USA, Canada, USA, USA, Finland, Kainuu Canada, British Columbia Finland, Kainuu Canada, British Columbia (epitype) et al. (2014) Velmala Canada, British Columbia Canada, British Columbia Finland, Koillismaa Finland, Koillismaa USA, Oregon Finland, Koillismaa Locality Canada, Prince Edward Island MAF-Lich. 20610 Canada, Prince Edward Island MAF-Lich. 20622 Canada, Prince Edward Island MAF-Lich. 20611 Canada, Prince Edward Island MAF-Lich. 20613 Canada, Prince Edward Island MAF-Lich. 20614 Canada, Quebec Canada, Quebec Canada, Quebec USA, Finland, Koillismaa Finland, Koillismaa Finland, Oulun Pohjanmaa Finland, Uusimaa Finland, Varsinais-Suomi Sicily Italy, Morocco, Rif Nord-Trøndelag Norway, Oppland Norway, Nord-Trøndelag Norway, Nord-Trøndelag Norway, Norway, Russia, Perm Russia, Perm Spain, Spain, Cáceres Spain, Lérida Spain, Lérida Spain, Navarra Spain, Navarra Sweden, Västerbotten Abs. = No substances detected; B. sp. B. vrangiana

B. pseudofuscescens

Table 1 (cont.) Table Taxon B. pikei (cont.)

C.G. Boluda et al.: Mismatch between phenotypes and genotypes in lichenized fungi 79 barcoding markers show that what were considered well delim- MATERIALS AND METHODS ited morphospecies are found admixed in a single lineage that may be interpreted as a single phylogenetic species. In such Sampling situations, many processes may be operative, including envi- We examined 142 specimens from 14 countries in Europe, ronmental plasticity (Boluda et al. 2016), hybridisation, ances- the Mediterranean Basin, and North and South America, rep- tral polymorphisms, incomplete lineage sorting (Joly et al. resenting 11 named morphospecies in Bryoria sect. Implexae 2009), limited value of neutral markers (Bekessy et al. 2003), (Table 1). Our dataset included 91 of the 97 specimens used or morphological variability mediated by low selective pressure, by Velmala et al. (2014) in their revision of B. sect. Implexae. genetic drift, or huge population sizes (Hartl & Clark 2007). In Newly obtained sequences are shown in bold in Table 1. Bryoria these cases, the use of additional markers, especially highly furcellata was used as outgroup to root the tree (Velmala et al. variable ones (e.g., microsatellites), may contribute to an ex- 2014). Names used in the analyses follow the species concepts planation of the underlying phenomena. adopted in Velmala et al. (2014). Chemical characters, mainly the production of polyketides, were accorded major importance in species delimitation in lichen- Morphology and chemistry forming fungi in the 1960s and 1970s (Hawksworth 1976, The newly studied specimens (Table 1) were examined mor- Lumbsch 1988). These compounds are formed by the fungal phologically under a Nikon SMZ-1000 dissecting microscope, partner, and that expression can differ according to the position and hand-cut sections studied with a Nikon Eclipse-80i com- in a thallus or in pure culture. For almost 50 years, chemical pound microscope equipped with bright field and differential products, generally linked to minor morphological differences, interference contrast (DIC). Habit photographs were taken have been used to circumscribe species in Bryoria (Hawksworth with a Nikon 105 mm f/2.8D AF Micro-Nikkor Lens coupled to 1972, Brodo & Hawksworth 1977, Myllys et al. 2011, Velmala et a Nikon D90 camera with daylight. Spot tests (K, C, and PD) al. 2014). The advent of molecular phylogenetics has enabled and TLC were carried out following Orange et al. (2010). Sol- such species concepts to be tested, and they have proved par- vent system C (200 ml toluene / 30 mL acetic acid) was used ticularly wanting in one group of species, those placed in Bryoria for TLC, with concentrated acetone extracts at 50 °C spotted sect. Implexae (Myllys et al. 2011, Velmala et al. 2014, Boluda onto silica gel 60 F254 aluminium sheets (Merck, Darmstadt, et al. 2015). Velmala et al. (2014) provided DNA sequence data Germany). Spotted sheets were dried for 10 min in an acetic for 11 species in the section, and with the exception of B. glabra, acid atmosphere to maximize resolution. Segments from the all the other species were intermixed in clades with diverse, and same lichen branch were used for both TLC and DNA extrac- not concordant, chemical and morphological features. Geneti- tion to avoid the possible risk of taking samples from mixed cally indistinguishable taxa (with the markers used), maintain collections. Morphological and thin layer chromatographic (TLC) distinctive phenotypes even when growing in physical contact analyses of the samples used in Velmala et al. (2014; Table 1) with one another (Velmala et al. 2014, Boluda et al. 2015), so were taken from that study. the variation cannot be attributed solely to ecological factors. A study on the morphospecies B. fuscescens in central Spain DNA dataset (Boluda et al. 2015) revealed specimens with the same nuclear The molecular dataset comprised DNA sequences and SSRs internal transcribed spacer sequence (nuITS) but different ex- frequencies. DNA extraction was performed with the DNeasy trolites (compounds formed on the surface of or excreted from Plant Mini Kit (Qiagen, Barcelona, Spain), following the manu- hyphae). Subsequent fieldwork across Europe has revealed facturer’s instructions. further combinations of extrolites, and also specimens sharing Eighteen fungal-specific microsatellites markers (Bi01, Bi02, characters of additional morphospecies. In order to under- Bi03, Bi04, Bi05, Bi06, Bi07, Bi08, Bi09, Bi10, Bi11, Bi12, Bi13, stand the evolutionary processes involved in B. fuscescens Bi14, Bi15, Bi16, Bi18 and Bi19) were amplified following Na- and related species we have adopted an integrative approach dyeina et al. (2014) using fluorescently labelled primers. Frag- including morphological, distributional, and chemical data to- ment lengths were determined on an ABI PRISM® 3130 Genetic gether with DNA sequences from three standard loci (Schoch et al. 2012), two newly tested loci, and eighteen microsatellite Analyser (Life Technologies, Carlsbad, CA, USA). Genotyping (SSRs) markers (Nadyeina et al. 2014). We then analysed was performed using GeneScan-500 LIZ as the internal size these datasets in a rigorous statistical framework to effectively standard and GeneMapper v. 3.7 (Applied Biosystems, Foster integrate an evolutionary perspective into a revised and defen- City, CA, USA). sible taxonomic treatment. These studies are reported here, For DNA sequencing, five loci were selected (Table 2), three and we anticipate that the experience gained in this group of commonly used as standard markers in fungi (ITS, IGS, and lichens will inform how other species complexes with similarly GAPDH), which were also used in Velmala et al. (2014), and discordant datasets can be addressed. two microsatellite flanking regions tested here for the first time

Table 2 Primer information used in Bryoria sect. Implexae.

Marker Description Primer forward (5’‒3’) Source Primer reverse (5’‒3’) Source

ITS Internal transcribed spacers ITS1-F: Gardes & Bruns (1993) ITS4: White et al. (1990) of the nuclear rDNA including the CTTGGTCATTTAGAGGAAGTAA TCCTCCGCTTATTGATATGC 5.8S region IGS Intergenic spacer of the IGS12b: Printzen & Ekman (2002) SSU72R: Gargas & Taylor (1992) nuclear rDNA AGTCTGTGGATTAGTGGCCG TTGCTTAAACTTAGACATG GAPDH Glyceraldehyde 3-phosphate Gpd1-LM: Myllys et al. (2002) Gpd2-LM: Myllys et al. (2002) dehydrogenase gene partial sequence ATTGGCCGCATCGTCTTCCGCAA CCACTCGTTGTCGTACCA FRBi15 Flanking region of Bryoria sect. FRBi15f: This paper FRBi15r: This paper Implexae microsatellite marker 15 GTCATAAGGGTATCAATCC TGAAAAGGTTTGGTGACTC FRBI16 Flanking region of Bryoria sect. FRBi16f: This paper FRBi16r: This paper Implexae microsatellite marker 16 CGAGGTTTCAGGAAAGGGAA AGGAAGTGATGTCGAGGT 80 Persoonia – Volume 42, 2019

(FRBi15 and FRBi16). Microsatellite flanking regions are vari- under accession nos TB2:S20007 (ITS, IGS, and GAPDH), able non-coding DNA fragments that can contain phylogeneti- TB2:S20005 (FRBi15), and TB2:S20004 (FRBi16). RDP v. 4 cal signal through a neutral molecular evolution (Zardoya et (Martin et al. 2010) was used to detect potential recombination al. 1996, Chatrou et al. 2009). To explore this possibility, the events, through the methods RDP (Martin & Rybicki 2000), flanking regions of the 18 microsatellite markers were checked GENECONV (Padidam et al. 1999), Chimaera (Posada & Cran- upstream and downstream in the 454 pyrosequencing contigs dall 2001), Maxchi (Maynard-Smith 1992), Bootscan (Gibbs et used for microsatellite searching in Nadyeina et al. (2014). al. 2000, Martin et al. 2005), SiScan (Weiller 1998, Gibbs et The variability of each region was assessed with the number al. 2000), PhylPro (Weiller 1998), and 3Seq (Boni et al. 2007). of variable sites in contigs supported by 2–16 copies. From the Partitionfinder (Lanfear et al. 2012) was used to detect possible 36 regions (two for each of the 18 microsatellites), the most intra-locus substitution model variability, resulting in the splitting variable flanking regions were in Bi15 and Bi16, and specific of the ITS region into ITS1, 5.8S, and ITS2, and coding each primers were designed for those loci (Table 2). codon position separately in GAPDH. Models of DNA sequence New DNA sequences (Table 1) were obtained using polymerase evolution for each locus partition were selected with jModeltest chain reactions (PCRs) as follows: a reaction mixture of 25 µL, v. 2.0 (Darriba et al. 2012), using the Akaike information criterion containing 12 µL sterile water, 9 µL JumpStartTM REDTaq (AIC, Akaike 1974). The best-fit model of evolution obtained ReadyMix PCR Reaction Mix (Sigma-Aldrich, St Louis, MI, was: ITS1 = TIM2, 5.8S = K80, ITS2 = TIM2ef + G, IGS = USA), 1.25 µL of each primer (forward and reverse) at 10 µM, TrN + I, GAPDH 1st position = TrN + I, GAPDH 2nd position = and a 1.5 µL DNA template. Cycling conditions for ITS, GAPDH, F81 + I, GAPDH 3th position = TPM3uf, FRBi15 = TPM3uf + I, FRBi15, and FRBi16 were 2 min at 94 °C; 35 cycles of 30 s FRBi16 = TPM3uf + G. To detect possible topological conflicts at 94 °C; 30 s at 56 °C; 2 min at 72 °C; and a final extension among loci, the CADM test (Legendre & Lapointe 2004, Camp- of 5 min at 72 °C. For IGS, the cycling process was: 2 min at bell et al. 2011) was performed using the function ‘CADM.global’ 94 °C; 15 cycles of 30 s at 94 °C, 30 s at 55 °C (decreasing implemented in the library ‘ape’ of R (Paradis et al. 2004). As 1 °C each cycle down to 40 °C), 2 min at 72 °C, then 35 cy- loci FRBi15 and FRBi16 were not congruent among them and cles of 30 s at 94 °C, 30 s at 55 °C; 90 s at 72 °C, and a final neither with the remaining loci, three alignments were used, extension of 5 min at 72 °C. PCR products were checked and resulting in three trees, one for each FRBi region and another for quantified on 1 % agarose gel stained with ethidium bromide the concatenated dataset including loci ITS, IGS and GAPDH. and cleaned using Exonuclease I and FastAP Thermosensitive For the concatenated matrix, specimens with more than one Alkaline Phosphatase (Thermo Fisher Scientific, Waltham, MA, missing locus were excluded. Datasets were analysed using USA) according to the manufacturer’s instructions. Sequencing maximum likelihood (ML) and Bayesian (B/MCMCMC) ap- was performed with labelling using BigDye Terminator v. 3.1 Kit proaches with gaps treated as missing data. (Applied Biosystems) as follows: 25 cycles of 20 s at 96 °C, 5 s For ML tree reconstruction, we used RAxML v. 8.2.10 (Stama- at 50 °C, and 2 min at 60 °C. PCR products were cleaned-up takis 2006) implemented in CIPRES Science Gateway (https:// with the BigDye XTerminator Purification Kit (Applied Biosys- www.phylo.org/; Miller et al. 2010) with the GTRGAMMA tems) according to the manufacturer’s instructions. Sequences model (Stamatakis 2006, 2014, Stamatakis et al. 2008). Sup- were obtained in an ABI PRISM 3130 Genetic Analyser (Life port values were assessed using the ‘rapid bootstrapping’ Technologies) and manually adjusted using DNA Workbench option with 1 000 replicates. For the Bayesian reconstruc- v. 6 (CLC bio, Aarhus, Denmark) and MEGA5 (Tamura et al. tion, MrBayes v. 3.2.1 (Ronquist & Huelsenbeck 2003) was 2011). Newly generated sequences were deposited in GenBank used. Two simultaneous runs with 10 M generations each, (Table 1 in bold). starting with a random tree and employing 12 simultaneous chains, were executed. Every 500th tree was saved to a file. Clustering methodologies Preliminary analysis resulted in an overestimation of branch lengths and to correct this we used the uniform compound Phenetic analyses Dirichlet prior brlenspr = unconstrained : gammadir (1, 1, 1, 1; Two presence/absence (1/0) matrices were constructed, one Zamora et al. 2015). We plotted the log-likelihood scores of for the extrolites detected by TLC, and another with morphology sample points against generations using Tracer v. 1.5 (Rambaut and geography data (Appendix 1). Morphological characters scored et al. 2014) and determined that stationarity had been achieved comprised those traditionally used to separate morphospecies when the log-likelihood values of the sample points reached in the group: an equilibrium and ESS values exceeded 200 (Huelsenbeck 1. pale/dark thallus colour; & Ronquist 2001). Posterior probabilities (PPs) were obtained 2. branching angles (acute/obtuse/mixed); from the 50 % majority rule consensus of sampled trees after 3. soralia (absent/fissural/tuberculate/both); and excluding the initial 25 % as burn-in. The phylogenetic tree was 4 pseudocyphellae (conspicuous/inconspicuous). drawn with FigTree v. 1.4 (Rambaut 2009). For distributions, Old World vs New World was used. The R package cluster (Maechler et al. 2013) was used to obtain STRUCTURE the dissimilarity matrix, and then the pvclust package (Suzuki & STRUCTURE v. 2.3.4 (Pritchard et al. 2000, Falush et al. 2003) Shimodaira 2006) was run to obtain a phenogram (Zamora et was run with the SSRs data matrix. Analysis was computed al. 2013). Multiscale bootstrap resampling with 10 000 bootstrap with 100 000 burn-in generations and 100 000 iterations using (bp) replicates was used to obtain approximately unbiased (au) a K value from 1 to 12 (i.e., the putative number of species p-values for branch supports. Groups were considered as sup- we may have) and 20 replicates for each K. To combine the ported when bp values exceeded 70 or au values exceeded 95. 20 runs of each K in a single result, CLUMMP v. 1.1.2 (Jakobs- son & Rosenberg 2007) was used and visualised replacing the Phylogenetic tree CLUMMP output values in a STRUCTURE output of the same Alignments for each locus were performed using MAFFT v. 7 K, and then plotted using the STRUCTURE software. To show (http://mafft.cbrc.jp/alignment/server/; Katoh & Standley 2013) the probability of each K value, STRUCTURE HARVESTER with the G-INS-i alignment algorithm, a ‘1PAM/K = 2’ scoring (Earl & Von Holdt 2012), with the ΔK method (Evanno et al. matrix, with an offset value of 0.1, and the remaining para 2005) was used, considering the most probable K the first one meters set as default. Alignments were deposited in TreeBASE that appears close to 0 in the output graphic. C.G. Boluda et al.: Mismatch between phenotypes and genotypes in lichenized fungi 81

Principal coordinate analysis adjust better to the data. The xml file was executed in BEAST Principal coordinate analysis (PCoA) was carried out with the with 250 M MCMC iterations, sampling every 10 000th iteration. SSRs length data in GenAlEx 6.5. The results of the three first Tracer v. 1.6 (Rambaut et al. 2014) was used to assess ESS axes were plotted in a three-axis graph using The Excel 3D values above 140. The resulting *BEAST species tree output Scatter Plot v. 2.1, in which the graphic can be moved in 3D to was then treated with SpeciesDelimitationAnalyzer (Jones et obtain a better understanding of how the plots are distributed in al. 2014), with a burn-in of 5 000 trees (20 % of the total gene the space. Since the projection of this 3D graph on a paper is rated), a collapse height of 0.001 (one fraction lower than in necessarily confusing, PCoA results were plotted on two differ- the *BEAST analysis) and a simcutoff value of 1 to ignore this ent 2D graphs showing axes 1 and 2, and 1 and 3, respectively. parameter, as according to sequence variability, we expected very similar putative species to emerge. The resulting similarity Haplotype network matrix was plotted with R v. 2.15.1 (R Core Team 2014) following the method of Jones et al. (2014). Haplotype network reconstruction was performed using TCS v. 1.2.1 (Clement et al. 2000) with the concatenated sequences Divergence time estimation matrix, excluding the outgroup, using gaps as missing data, and a 95 % connection limit. Specimens differing only by missing or Two divergence time estimations were performed, one only with ambiguous characters were not counted as haplotypes. the ITS region and a defined substitution rate, and the other with the concatenated data matrix of ITS, IGS, and GAPDH Species delimitation analyses loci. A rate of 3.30 × 10‒9 s∙s‒1∙yr‒1 for the ITS region as a In order to examine species delimitation, four computational whole was used, with a GTR + G + I substitution model (Leavitt et al. 2012a, b). In the concatenated matrix analysis, as no approaches not requiring prior hypothesis of a putative number previous literature on substitution rates for IGS and GAPDH in of species were used: lichen-forming fungi is available, these were set as estimated 1. Automatic Barcode Gap Discovery ABGD (Puillandre et al. in the Bayesian phylogenetic analysis. A *BEAST analysis was 2011) based on barcode gaps using genetic distances; executed, using a relaxed clock model (uncorrelated lognormal), 2. Poisson Tree Processes PTP (Zhang et al. 2013), based a birth-death model prior for the node heights and unlinked on gene trees; substitution models, clocks and trees for each partition. Clades 3. The Generalized Mixed Yule coalescent approach GMYC, G, Ko, NA, and WD were selected as potential species, forc- which combines a coalescent model of intraspecific branching them to remain monophyletic (Fig. 6). No calibration points ing with a Yule model for interspecific branching (Pons et could be used, as no fossils or previous dating of this species al. 2006, Monaghan et al. 2009); and complex are available. To avoid stochastic events, two inde- 4 DISSECT (Jones et al. 2014) based on the multispecies pendent analyses were run, each with 200 million generations, coalescent model for species delimitation. sampling each 5 000 trees, and discarding the first 10 000 trees ABGD and PTP were carried out using the online servers http:// (25 %) as burn-in. Tracer v. 1.6 (Rambaut et al. 2014) was used wwwabi.snv.jussieu.fr/public/abgd/ and http://species.h-its. to ensure ESS parameter values above 115 in the concatenated org, respectively. GMYC was analysed with the gmyc func- matrix, and 185 for the ITS analysis. Different priors were tion in the SPLITS package in R (v. 2.10, www.cran.r-project. tested but no higher ESS values could be obtained, which we org), employing the single (GMYCs) and multiple (GMYCm) suspect was due to the very similar sequences, and the uncer- threshold methods. Because GMYC needs a strictly ultramet- tain topology of the backbone connecting the groups Ko, NA, ric and bifurcating tree with no zero branch lengths, identical and WD. The two runs performed for each input were merged sequences were deleted and an ultrametric tree was generated with logcombiner v. 1.8.2, and the resulting trees merged in a using BEAST v. 1.8.2 software (Drumond et al. 2012), with the consensus tree using TreeAnnotator v. 1.8.2 (Drumond et al. evolutionary models explained in the Bayesian phylogenetic 2012). FigTree v. 1.4 (Rambaut 2009) was used to display the reconstruction. A run of 100 M iterations logging every 1 000th ITS and the concatenated dated species trees. iteration was conducted. Consensus tree was generated with TreeAnotator v. 1.8.2 after discarding the initial 10 % trees as Demography burn-in. ESS values above 200 were ensured using Tracer Changes in population sizes through time were estimated us- v. 1.6 (Rambaut et al. 2014). ing the Bayesian skyline analysis (Drumond et al. 2005) with DISSECT analysis was implemented in STARBEAST (*BEAST, BEAST. Only clades Ko, NA, and WD, isolated and merged, Drumond et al. 2012) using the concatenated DNA matrix after were studied, as they show a clock-like tree topology and ad- removing identical sequences and following the instructions of equate sampling sizes. Jones et al. (2014). First, we used BEAUti (Drumond et al. 2012) Following the methods used for divergence time estimation to produce the xml file, with every individual encoded as if it was analysis, the demography analyses were run using the ITS a separate species. Sites, clocks and trees were released as region without partitioning, with the GTR + G + I model of unlinked. Nucleotide substitution models and other parameters nucleotide substitution and a substitution rate of 3.30 × 10‒9 (as in the Bayesian analysis, see above), were encoded using s∙s‒1∙yr‒1, and with a strict molecular clock model (Leavitt et BEAUti if possible, or manually entered. For the ITS locus, a al. 2012a, b). Additionally, the same analysis was repeated with substitution rate of 0.0033 substitutions per site per million years the concatenated data matrix using the ITS substitution rate, was introduced (Leavitt et al. 2012a, b), setting other loci as estimating the other loci rates with a relaxed clock model and estimated with a lognormal relaxed clock. A birth-death-collapse using the nucleotide substitution models for IGS and GAPDH prior that controlled the minimal split heights for the putative explained in the Bayesian phylogenetic reconstruction. Four resulting species was manually added to the xml file. This prior independent runs for each input were processed with 50 M contained the parameters CollapseHeight (ε) with a value of MCMC generations, sampling parameter values every 5 000th 0.0001 and CollapseWeight (ω), set as estimated using a Beta generation, using the Bayesian Skyline tree prior model, six distribution with values 10 and 1.5. Selected parameters provide discreet changes in population size and the linear growth op- a highest probability density around 4‒5 clusters, the most tion. ESS values were checked with Tracer v. 1.5 (Rambaut et probable number of taxa meriting separation according to other al. 2014), and the two best of the four runs were combined, analyses performed for this paper. However, this prior is diffuse obtaining values usually above 200, with some exceptions with and allows to obtain a different number of putative taxa if they a lower limit of 100. Skyline plots were drawn with Tracer v. 1.5. 82 Persoonia – Volume 42, 2019

To support the Bayesian skyline test, a neutrality test was per- taxa, as well as specimens producing extrolites characteristic formed to infer if populations are in mutation-drift equilibrium. of different taxa in a single thallus. Thin-layer chromatography Tajima’s D (Tajima 1989) and Fu’s Fs (Fu 1997) were calculated (TLC) revealed seven different extrolites: alectorialic, barbatolic, with DnaSP v. 5.10 (Librado & Rozas 2009). A significantly fumarprotocetraric, gyrophoric, norstictic, and psoromic acids, positive D is interpreted as a diversifying selection or a recent atranorin, and sometimes also related substances such as bottleneck, whereas a negative significant D shows purifying chloroatranorin, protocetraric, or connorstictic acids. Atranorin, selection or a recent expansion. If D is not significantly different a typical accessory substance in the genus Bryoria, was not from 0, a mutation-drift equilibrium may be occurring. Fu’s Fs used in the posterior analyses because it appears in trace can be interpreted in the same way. amounts in many samples and is often difficult to unequivocally discern if it is present or absent by TLC alone. The chemical presence/absence matrix resulted in the pheno- RESULTS gram shown in Appendix 2a. The matrix included specimens with as many as four extrolites, something not previously Morphological and chemical clustering reported in the complex. Chemical characters were separated The wide geographical range of collections revealed a combi- into two main groups: nation of characters not previously reported in the Bryoria fus- 1. specimens that contain benzyldepsides (i.e., alectorialic cescens complex, especially those from the previously less- and barbatolic acids), substances traditionally used to se studied Mediterranean Basin. Specimens with intermediate parate B. capillaris and B. pikei from other species in the morphologies amongst traditionally accepted species were rec- complex; and ognized, and the application of species names according to the 2. specimens without benzyldepsides. current taxonomy was ambiguous. Individuals connecting the The latter were clustered in two well-supported groups, one with phenotypes and chemotypes of the taxa currently recognized as fumarprotocetraric acid as the main substance, and the other Bryoria fuscescens, B. implexa, B. kuemmerleana, and B. vran- without it (including specimens with no detectable substances). giana were not rare in the Mediterranean Basin. For example, If the structural relationships of the compounds were encoded in chemotypes thought to be diagnostic for a particular taxon the presence/absence matrix (benzyldepsides vs depsidones), were detected in specimens morphologically belonging to other the same clustering was obtained.

Fig. 1 Phylogenetic relationships in Bryoria sect. Implexae based on FRBi15 and FRBi16 loci. Tree topology depicts the result of the Bayesian Markov chain Monte Carlo (B/MCMC) analysis. Posterior probabilities and bootstrap analysis for the supported nodes (≥ 0.95 and ≥ 70 %) are indicated at the main nodes. Lines connecting clades indicate putative recombination events, with main parents (continuous lines) and minor parents (discontinuous lines). Because the clade insertion in the trees is influenced by the recombination, clades with recombination are depicted with a discontinuous branch line. Note that clades with recombination appear as sister or close to the main parent but tending to be deviated towards the minor parent. ― The coloured bar corresponds to the SSR genepool from Fig. 2, with specimens intermediate between two or more genepools in grey. The clades obtained, although well-supported, do not follow any evident geographic, morphologic, chemical, or microsatellite pattern. C.G. Boluda et al.: Mismatch between phenotypes and genotypes in lichenized fungi 83

The analysis of combined morphological, geographical, and 2. FRBi15 with 93 individuals contained 569 unambiguously chemical characters resulted in the phenogram in Appendix 2b. aligned nucleotide position characters, with 44 Pi sites; and Only terminal branches were supported, including few mono- 3. FRBi16 with 80 individuals had 632 unambiguously aligned phyletic morphospecies, although not well isolated from others. nucleotide position characters, with 160 Pi sites. Neither accepted morphospecies nor an unequivocal number No evidence of recombination events was detected in the of phenotypic groups could be recognized. This ambiguity was concatenated matrix. The resulting tree (Fig. 6) had four well- largely attributable to phenotypically intermediate specimens, supported main clades, G (Glabra, yellow), Ko (Kockiana, mainly from the Mediterranean Basin, and also by the presence magenta), NA (North American, blue), and WD (Widely Dis- of some shared characters amongst the morphospecies, such tributed, red + green + brown). Clade G included only material as the presence/absence of soralia, pseudocyphellae, extrolite of B. glabra, appearing as an isolated taxon sister to the other composition, and thallus colour. three clades, which showed an uncertain topology between them. Clade Ko included material named B. kockiana and two Phylogenetic tree unidentiﬁed specimens, all collected in Alaska (USA). Clade NA Due to the topological conflict between loci, three DNA matrices comprised the previously recognized North American morphos- were used to generate three phylogenetic trees: pecies group (the ‘North American endemic species’, Velmala 1. a concatenated matrix including ITS, IGS, and GAPDH et al. 2014) named as B. friabilis, B. inactiva, B. pikei, and B. with 134 individuals consisting of 1 774 unambiguously pseudofuscescens. While these species were mixed in the tree, aligned nucleotide position characters, with 83 parsimony the group as a whole was resolved as monophyletic. The WD informative (Pi) sites; clade included specimens widely distributed but mainly from

Fig. 2 Bayesian inference of population structuring using STRUCTURE v. 2.3.4 (Pritchard et al. 2000, Falush et al. 2003) and nine microsatellite loci in Bryoria sect. Implexae. – Left. Results from the hypothesis of 2–6 clusters. Vertical bars represent specimen assignment probability into a genetic cluster depicted by the colours. Morphospecies names given to the specimens appear at the top. – Right. Detailed columns of the K = 6 hypothesis, the numbers representing the specimens shown in Table 1 to provide a better understanding of the components of each individual. ― G = Glabra clade; Ko = Kockiana clade; NA = North American clade; WD = Widely Distributed clade; * = Bryoria pikei specimen 49. 84 Persoonia – Volume 42, 2019

Europe (the ‘European and globally distributed species’ group, STRUCTURE clustering Velmala et al. 2014) under the names B. capillaris, B. fusces- Of the 18 microsatellite markers, the nine that showed more cens (syn. B. chalybeiformis and B. lanestris), B. implexa, B. than 95 % successful ampliﬁcation across the samples were kuemmerleana, and B. vrangiana. None of these previously used (number of haplotypes shown in brackets, Appendix 3): recognised species formed a monophyletic group. Bi01 (17), Bi03 (6), Bi04 (8), Bi05 (5), Bi10 (8), Bi11 (9), Bi12 The phylograms produced using the FRBi15 and FRBi16 mark- (12), Bi14 (6), and Bi19 (5). We allowed a maximum of three ers had a different tree topology, not congruent among them missing loci per specimen, a value reached only in seven or with that from the preceding concatenated dataset. In the samples. STRUCTURE was allowed to run to K = 12, but from FRBi15 reconstruction (Fig. 1), B. glabra was not represented K = 6 the clustering process started to be uninformative (Fig. 2). due to the lack of primer annealing in the PCR process, and The likelihood results of the ΔK analysis (Evanno et al. 2005) the tree could not be rooted. Several well-supported groups indicated three as the most probable number of clusters (likeli- were produced, but did not follow any evident geographic, mor- hood = ‒1232, ΔK = 2.2), the clades G, NA, and WD (Fig. 2, phological, chemical, or SSR frequency pattern. Bryoria pikei K = 3). Clade Ko, which appeared isolated in the concatenated L376 had a sequence with a putative recombination fragment phylogenetic tree (Fig. 6), could not be accepted as distinct with the B. vrangiana S10 clade in c. 50 % of the total length. under a K = 3 hypothesis. However, B. glabra, a morphologi- This insertion placed the specimen out of the main parental cally delimited taxon, was not isolated at ﬁrst in STRUCTURE. group and it appears as its sister. Although marker FRBi16 (Fig. This could be attributable to the clustering algorithms being 1) also produced a well-resolved tree with supported nodes, influenced by unbalanced sampling sizes, masking clade Ko, the clades do not show any phenotypic and/or geographic which appeared isolated at K = 6. From K = 4 to K = 6, the structure. In this tree reconstruction, B. glabra was an isolated new groups appeared mainly inside the WD clade, showing taxon and served to root the tree. In FRBi16 sequences, many that the samples from Europe were much more diverse than putative recombination events were detected, suggesting a those from North America. Indeed, the NA clade was not split reticulate evolution. In both trees in Fig. 1, clade Ko (magenta) into subgroups even at K = 10. Apart from B. glabra, no other was recovered as monophyletic, but embedded between other named morphospecies formed an exclusive cluster even at named morphospecies. high K values.

Fig. 3 Principal Coordinate Analysis (PCoA) of microsatellite data in Bryoria sect. Implexae. Species names according to Velmala et al. 2014 (shape and colours) and STRUCTURE clusters (colours) for each specimen are represented in the three main coordinates. Note that the Ko clade does not appear isolated from the NA clade in any coordinate axis. ― G = Glabra clade; Ko = Kockiana clade; NA = North American clade; WDr = Widely Distributed red clade; WDg = Widely Distributed green clade.

Table 3 Species delimitation analysis results for loci ITS, IGS, GAPDH and the concatenated data matrix in Bryoria sect. Implexae. Brackets indicate groups predicted as conspeciﬁc. ― G = Glabra clade; Ko = Kockiana clade; NA = North American clade; WD = Wide Distributed clade; WDr = Wide Distributed red clade; WDg = Wide Distributed green clade; pik5 = Specimen Bryoria pikei 5.

Method ITS IGS GAPDH Concatenated

ABGD 2 = G + (Ko, NA, WD) 2 = G + (Ko, NA, WD) 4 = G + Ko + NA + WD 4 = G + Ko + NA + WD PTP 2 = G + (Ko, NA, WD) 2 = G + (Ko, NA, WD) 4 = G + Ko + NA + WD 5 = G + Ko + NA + WDr + WDg GMYCs 4 = G + (Ko, NA, WDg) + WDr + WDr 3 = G + (Ko, WD) + NA 4 = G + Ko + NA + WD 6 = G + Ko + NA + pik5 + WDr + WDg GMYCm 4 = G + (Ko, NA, WDg) + WDr + WDr 4 = G + (Ko, WD) + NA1 + NA2 4 = G + Ko + NA + WD 5 = G + Ko + NA + WDr + WDg DISSECT ‒ ‒ ‒ 5 = G + Ko + NA + pik5 + WD C.G. Boluda et al.: Mismatch between phenotypes and genotypes in lichenized fungi 85

Principal coordinates analysis (PCoA) whereas the coalescence methods (especially GMYC) the The PCoA analysis has a three-dimensional output represented largest. Analyses also revealed the contribution of each locus here in two graphs, one comparing axis 1 against 2, and the to the postulated species delimitation, GAPDH being the most other 1 against 3 (Fig. 3). The information percentage of each informative and constant marker. DISSECT analysis (Fig. 4) axis was 44.47 %, 15.06 %, and 14.44 %, respectively. Clade predicted five species corresponding to G, Ko, NA, and WD G (Fig. 3, yellow) appeared isolated, whereas clade Ko (Fig. clades, and specimen B. pikei 5. Although the GMYC analysis 3, magenta) was admixed with NA clade (Fig. 3, blue), form- also showed the B. pikei 5 specimen as a separate species, it ing a single cluster. Clade WD was isolated from the others, was grouped in the NA clade in the other analyses. DISSECT but divided into two clusters, one corresponding to the red showed two internal greyish square groups in WD, but they did and brown groups in the K = 6 STRUCTURE output (Fig. 2), not correspond exactly to the WDr and WDg groups in Fig. 2, 3 and one for the green group. Apart from B. glabra, none of the (STRUCTURE and PCoA analyses). currently accepted morphospecies formed a defined group. Four reasonably isolated clusters could be distinguished, cor- Node dates and demographic history responding to the groups G, WDr (Widely Distributed, Fig. 3, The calibrated maximum clade credibility chronogram for the red), WDg (Widely Distributed, Fig. 3, green), and Ko together concatenated data matrix is shown in Fig. 6. As only the ITS with NA forming a single cluster. mutation rate is estimated in previous studies (Leavitt et al. 2012a, b), a second chronogram was prepared using this locus Haplotype network alone. Results from this analysis have to be treated with cau- The haplotype network of the concatenated data matrix, cod- tion, as the species tree is not strictly clock-like (B. glabra has a ing gaps as missing data, produced 39 haplotypes. Bryoria shorter branch), and the ITS mutation rate has been taken from glabra specimens (Appendix 4, yellow) formed two haplotypes Melanohalea, a lichen-forming genus in the same family. Both not connected with other members of the network, indicat- analyses produced similar values, and the divergence of the ing genetic isolation of this species. One of the haplotypes B. glabra lineage was estimated at 6.9 Mya (95 % HPD = 3.5– was composed exclusively of South American specimens, 10.8) in the concatenated matrix analysis, and 6.5 Mya (95 % whereas the other contained European, North American, and HPD = 2.2–11.4) in the ITS data alone. The Ko, NA, and WD South American samples. Clade Ko (Appendix 4, magenta) fell split was estimated at 1.0 Mya (95 % HPD = 0.3–2.2) from the into two haplotypes, one including specimens with psoromic concatenated matrix and 0.6 Mya (95 % HPD = 0.2–1.5) from acid and identified as B. kockiana, and the other clustering the ITS data alone. unidentified samples with no substances detected. This group Bayesian Skyline Plots (Fig. 5, left) indicate a recent population was connected to the NA clade (Appendix 4, blue) by a long increase in the NA and WD clades. However, the sequences branch with 13 mutation steps. The NA clade was separated contained few informative mutations and the deepest coales- by nine mutations from the WD clade (Appendix 4, green, red, cence was reached in around 700 000 yr, with no population and brown). The WD green, red, and brown groups split by changes detectable further back from this period. Tests of neu- STRUCTURE (Fig. 2) formed a unique cluster. Four isolated trality (Fig. 5, right) are commonly used to support inferences clusters could be distinguished, corresponding to the groups from Bayesian Skyline Plots. As indicated by non-significant G, Ko, NA, and WD. Tajima’s D and Fs results, all sampled groups seem in mutation- Species delimitation programs drift equilibrium, with the exception of the GAPDH locus of the NA clade which had a significant negative D value (Fig. 5 bold). The ABGD, PTP, GMYC, and DISSECT programs (Table 3) This could indicate a recent expansion or ‘purifying’ selection, use different algorithms, and consequently different numbers as seen in the concatenated Bayesian Skyline analysis, but of putative species may be predicted. The genetic distance other loci did not support this hypothesis. method (ABGD) gave the smallest number of putative species, Markers ITS, IGS, and GAPDH indicate population stability over the recent past for clades NA and WD, with putative even more recent small population expansions. Due to the low variability of the loci, and the putative loss of demographic signals, this hypothesis is not confirmed by this analysis.

Integrated assessment of datasets Depending on the analysis, different numbers of putative species were suggested, ranging from four to six (Table 4). All analyses, however, conﬁrmed that the combination of morphological and chemical characters generally used for species cir- cumscription in the complex was inadequate. GAPDH, despite its low variability, was the only marker tested that supported species-rank assignations for the clades G, Ko, NA, and WD (Table 3). ITS, one of the most used loci for DNA barcoding in lichen-forming fungi, did not unambiguously distinguish those clades. The new markers FRBi15 and FRBi16, despite their higher variability, showed inconclusive results and putative recombination events. The microsatellite data (Fig. 2) supported the DNA sequences results and reflected internal variability not Fig. 4 Similarity matrix from DISSECT analysis performed after clone cor- revealed in our sequence data, showing that the WD cluster rection in Bryoria sect. Implexae. Squares represent posterior probability was much more diverse than NA, which had a particularly low (white = 0, black = 1) of pairs of specimens to belong to the same species. diversity. Resulting major groups are delimited by lines, which indicate the clade on the collapsed phylogenetic tree. 86 Persoonia – Volume 42, 2019

Fig. 5 Bryoria sect. Implexae. – Left. Bayesian Skyline Plots for each clade predicted by the ITS marker and the concatenated loci matrix. The X-axis of each graph represent time (in Myr), and the Y-axis represents the value for the log of the effective population size as relative changes, because generation times in Bryoria species are unknown. Grey shadows indicate the upper and lower 95 % credible intervals. – Right. Results from neutrality tests for each marker and clade, indicating (in bold) any statistically signiﬁcant deviation from neutrality. ― h = number of haplotypes; Fs = Fu’s Fs; D = Tajima’s D.

Table 4 Summary of the number of putative species suggested by the different methods used for each dataset in Bryoria sect. Implexae.

Method Data Figure / reference Number of putative species

Traditional concept DNA sequences and phenotypic Velmala et al. (2014) 12 Chemical Phenotypic Appendix 2 Left c. 4 Morpho-chemical Phenotypic Appendix 2 Right Not conclusive Phylogeny DNA sequences of ITS, IGS, and GAPDH Fig. 6 4 = G + Ko + NA + WD Phylogeny DNA sequences of FRBi15 Fig. 1 Not conclusive Phylogeny DNA sequences of FRBi16 Fig. 1 Not conclusive STRUCTURE Microsatellites Fig. 2 5 = G + Ko + NA + WDr + WDg PCoA Microsatellites Fig. 3 4 = G + (Ko, NA) + WDr + WDg Haplotype Network DNA sequences of ITS, IGS, and GAPDH Appendix 4 4 = G + Ko + NA + WD ABGD DNA sequences of ITS, IGS, and GAPDH Table 3 4 = G + Ko + NA + WD PTP DNA sequences of ITS, IGS, and GAPDH Table 3 5 = G + Ko + NA + WDr + WDg GMYCs DNA sequences of ITS, IGS, and GAPDH Table 3 6 = G + Ko + NA + pik5 + WDr + WDg GMYCm DNA sequences of ITS, IGS, and GAPDH Table 3 5 = G + Ko + NA + WDr + WDg DISSECT DNA sequences of ITS, IGS, and GAPDH Fig. 4 5 = G + Ko + NA + pik5 + WD

TAXONOMY sent or present, tuberculate or fissural, white to dark. Isidia or isidioid spinules absent. Apothecia mainly absent, if present, Bryoria sect. Implexae (Gyeln.) Brodo & D. Hawksw., Opera usually afunctional. Chemistry varied, with no detectable or with Bot. 42: 114. 1977 one or a combination of major substances, including alectorialic, barbatolic, connorstictic, fumarprotocetraric, gyrophoric, Basionym. Bryopogon sect. Implexae Gyeln., Feddes Repert. Spec. Nov. norstictic, protocetraric, psoromic and possibly salazinic acids, Regni Veg. 38: 223, 238. 1935. atranorin, and chloroatranorin. Photobiont Trebouxia ‘hypogym- Type species. Bryoria implexa (Hoffm.) Brodo & D. Hawksw. 1977. ≡ niae’ (Lindgren et al. 2014). Usnea [unranked] implexa Hoffm. 1796. = Bryoria fuscescens (Gyeln.) Brodo & D. Hawksw. 1977; but see below. Notes ― Most species included in Brodo & Hawksworth (1977) under Bryoria sect. Implexae were transferred to other Species with a fruticose, hair-like, subpendent to mainly pen- sections in Myllys et al. (2011). In the light of our results (but dent thallus, lateral spinules or spinulose branches absent, see the Discussion later), Bryoria sect. Implexae includes the whitish grey to brown or black, often paler in the basal parts. four species treated below. Comments on particular morpho- Angles between branches variable, acute to obtuse or even logical or chemical traits that may be helpful for distinguishing rounded. Pseudocyphellae absent or present, then frequently these taxa are given under each species. Nevertheless, nearly inconspicuous, ± fusiform, concolorous or whitish. Soralia ab- all cited characters are shared by different taxa, so they can C.G. Boluda et al.: Mismatch between phenotypes and genotypes in lichenized fungi 87 be interpreted as ‘cryptic’. The species names adopted here name in the complex, is well established, the most widely used* are epitypified by sequenced material here in order to fix their and is already conserved over two earlier species names in the identities at the molecular level. This epitypification is essential complex. In addition, all the other names have been associated to fix the application of these names as no DNA sequences are with particular morphotypes or chemotypes since the 1970s, available and cannot be obtained from old type material of most and so their use might be mistaken as applying to a taxon with species names. The old types cannot therefore be critically those particular traits. identified for purposes of the precise application of the names If the proposal for rejection of the previously mentioned compet- and so epitypes may be designated (Turland et al. 2018: Art. ing synonyms is not accepted, the principle of priority would 9.9 and Ex 9). rule the use of the earliest and not already rejected, validly As molecular data are necessary for unambiguous species level published name at the species rank, i.e., Usnea implexa (and identification in the taxonomy proposed here, we recommend then the combination Bryoria implexa), which would require using the collective ‘Bryoria fuscescens complex’ when referring epitypification by sequenced material in order to fix the precise to material lacking molecular data. application of that name. The species was first described from Germany but with no named locality, and neotypified by an Bryoria fuscescens (Gyeln.) Brodo & D. Hawksw., Opera unlocalised and undated specimen in Hoffmann’s herbarium Bot. 42: 83. 1977 in Moscow (Hoffmann 8569, MW) which may be part of the original material from Germany or have been collected later Basionym. Alectoria fuscescens Gyeln., Nytt Mag. Naturvidensk. 70: 55. and perhaps in Russia (Hawksworth 1969a). As the neotype 1932, nom. cons. (cf. Hawksworth & Jørgensen 2013). contained psoromic acid, and the epithet has therefore been Synonyms. Lichen chalybeiformis L., Sp. Pl. 2: 1155. 1753, (nom. cons.) applied to that chemotype, a potential sequenced epitype nom. rej. against Bryoria fuscescens (cf. Hawksworth & Jørgensen 2013). should represent that chemotype and ideally also have been Bryoria chalybeiformis (L.) Brodo & D. Hawksw., Opera Bot. 42: 81. 1977. collected in Germany. No such specimen was available to us Usnea [unranked] implexa Hoffm., Deutschl. Fl., Zweiter Teil: 134. 1796. Bryoria implexa (Hoffm.) Brodo & D. Hawksw., Opera Bot. 42: 121. 1977. during this study. Parmelia jubata β. [var.] capillaris Ach., Methodus, Sectio post.: 273. Distribution ― Widely distributed, known from cool temper- 1803. ate to boreal and arctic areas of Europe, Asia, North America, Bryoria capillaris (Ach.) Brodo & D. Hawksw., Opera Bot. 42: 115. 1977. and Africa. There are also reports from Antarctica, Oceania, Alectoria jubata ζ. [var.] lanestris Ach., Lichenogr. Universalis: 593. 1810. and South America, but material from those regions has not Bryoria lanestris (Ach.) Brodo & D. Hawksw., Opera Bot. 70: 88 1977. yet been studied molecularly and so we cannot confirm that Alectoria kuemmerleana Gyeln., Nytt Mag. Naturvidensk. 70: 49. 1932. they belong to this complex. Bryoria kuemmerleana (Gyeln.) Brodo & D. Hawksw., Opera Bot. 42: 155. 1977. Notes ― Bryoria fuscescens is highly variable in morphology Alectoria prolixa var. subcana Nyl. ex Stizenb., Ann. Naturhist. Mus. and chemistry, and many of the analysed specimens develop Wien 7: 129. 1892, nom. rej. against Bryoria fuscescens (cf. Hawksworth & soralia. Further, atranorin, which is not normally detectable Jørgensen 2013). in the other three species accepted here, is frequently found Bryoria subcana (Nyl. ex Stizenb.) Brodo & D. Hawksw., Opera Bot. 42: 91. 1977. in concentrated extracts from both sorediate and esorediate Alectoria vrangiana Gyeln., Magyar Bot. Lapok 31: 46. 1932. morphs. Bryoria vrangiana (Gyeln.) Brodo & D. Hawksw., Opera Bot. 42: 97. 1977.

Type specimens. Finland, Tavastia austr., Hollola, ad truncos Pini locis Bryoria glabra (Motyka) Brodo & D. Hawksw., Opera Bot. 42: apricioribus in silva, Sept. 1882, J.P. Norrlin (Norrlin, Herb. Lich. Fenn. No. 86. 1977 46) (BP 33947 – lectotype designated by Hawksworth 1972: 217). ― Finland, Basionym. Alectoria glabra Motyka, Fragm. Florist. Geobot. 6: 448. 1960. Etelä-Savo, Savitaipale, 600 m NW of Mustapää, 61, N1721° E27,6900°, 2005, L. Myllys 464 (HA.H9209715 (L139)) – epitype designated here, Myco Type specimens. USA, Washington, Olympic Peninsula, Clallam Co., Hur- Bank MBT381730. ricane Ridge, 5800 ft, on trunk of Abies lasiocarpa, 24 July 1950, B.I. Brown & W.C. Muenscher 129 (US – holotype). ― USA, Alaska, Mainland, Valley Nomenclature ― A large number of species rank names between the Bucher and Gilkey Glaciers, southern end of subalpine valley, belong to this group, and are synonymised, but these have not on east side of creek running through valley, subalpine forest, N58°47'20.12" been epitypified with sequenced material. Further information W134°30'0.10", 773 m elevation, on Tsuga mertensiana twigs, 4 Aug. 2011, on synonyms and type materials can be seen in Hawksworth K.L. Dillman 4Aug11:1 (UBC (L406)) – epitype designated here, MycoBank (1972), Brodo & Hawksworth (1977) and Velmala et al. (2014). MBT381731. Although no samples of Bryoria austromontana have been Distribution ― Known from northern Europe (Scandinavia), studied, the published description and images (Jørgensen & and North and South America. In North America, it is most Galloway 1983) suggest this taxon also belong here. abundant in the Pacific North-West. The earliest species rank epithets amongst these are chalybeiformis dating from 1753 (Lichen chalybeiformis), and implexa Notes ― Distinguishing features in well-developed speci- dating from 1799 (Usnea implexa). The former has been re- mens are the brownish thallus with a regular branching pat- jected against Bryoria fuscescens, but not against other species tern, generally with obtuse and rounded angles between the names apart from B. subcana (Hawksworth & Jørgensen 2013). branches, and broad oval and usually white soralia. It is, how- A proposal to add the four earlier names Alectoria capillaris, ever, difficult to separate poorly developed or small specimens Usnea implexa, A. kuemmerleana, and A. lanestris to the two conclusively, so molecular sequences are recommended for against which Alectoria fuscescens is already conserved is unambiguous identifications. Only fumarprotocetraric and pro- being prepared separately. Protection against A. vrangiana is tocetraric acids have been detected in this species, and these not required as it appeared in the same work as A. fuscescens. are characteristically produced in the soralia. While the proposal is under discussion, the name B. fuscescens The Alaskan specimen is selected as the epitype here as se- should be adopted in accordance with Rec. 14A.1 (Turland et quences are available from all loci, whereas the material we al. 2018). * Hits obtained for these names in Google and Google Scholar respectively We refrained from epitypifying and taking up any of the earlier on 12 April 2018 were: B. fuscescens (12 300 and 1 620), B. capillaris and potentially competing names by epitypification primarily as (10 300 and 999), B. implexa (6 020 and 447), B. kuemmerleana (226 and the name B. fuscescens is the most commonly used species 20), B. lanestris (6 740 and 185), and B. vrangiana (532 and 61). 88 Persoonia – Volume 42, 2019 have sequenced from Washington state (type locality) only has Sexual structures are of major importance in species identifi data on the ITS locus. cation in fungi, but here the rarity of apothecial production has hampered their study in most Bryoria species. Any such features found would in any case be of limited practical value Bryoria kockiana Velmala, Myllys & Goward in Velmala et al., in identification as nearly all samples lack apothecia. Extrolite Ann. Bot. Fenn. 51: 361. 2014 composition has been accorded a major role in species delimi- Type specimen. USA, Alaska, Fairbanks, North Star Borough, 26 July tation in the complex, but many of the substances that were 2011, D. Nossov 20019-1 (UBC (L394) – holotype). considered to be of diagnostic value are biosynthetically closely Distribution ― Known only from Alaska (USA) and British related, being produced by the same gene cluster (pks genes; Columbia (Canada), on conifer branches. Keller & Hohn 1997), and may be environmentally influenced (Myllys et al. 2016, Lutsak et al. 2017). Notes ― Few specimens of this species have so far been Integrative taxonomy, rather than phylogenies based only on studied, and these are characterised by the absence of any neutral markers, are increasingly being used to resolve com- whitish grey tone in the thallus, the lack of soralia, and greyish plex taxonomical groups (e.g., Dayrat 2005, Will et al. 2005, to brown branches with conspicuous, white to concolourous, Lumley & Sperling 2011, Zamora et al. 2015, Caparrós et al. broad, elongate-fusiform, sometimes slightly raised pseudo- 2016). Microsatellites are also now widely used in intraspecific cyphellae. It lacks TLC-detectable substances or produces population studies because of their high variability (Widmer et psoromic acid. The not validly published designation Alectoria al. 2012, Dal Grande et al. 2014), and in species complexes krogii D. Hawksw. 1972 may be synonymised here. with diffuse genetic barriers, microsatellite data can improve DNA sequence resolution (Lumley & Sperling 2011, Vanhaecke Bryoria pseudofuscescens (Gyeln.) Brodo & D. Hawksw., et al. 2012). It is generally assumed that DNA sequence data Opera Bot. 42: 127. 1977 reflect the evolution of the species, but these data only reflect Basionym. Alectoria pseudofuscescens Gyeln., Ann. Hist.-Nat. Mus. Natl. the history of the studied loci, which may sometimes be different Hung. 28: 283. 1934, and Rev. Bryol. Lichénol. 7: 51. 1934. from the species history overall (Nichols 2001). In this case, Synonyms. Bryoria friabilis Brodo & D. Hawksw., Opera Bot 42: 118. we demonstrated that traditionally used loci (ITS, IGS, and 1977. GAPDH) and microsatellite data reveal similar clades, whereas Bryoria pikei Brodo & D. Hawksw., Opera Bot 42: 125. 1977. other intergenic loci (FRBi15 and FRBi16) produced discrepant Bryoria inactiva Goward et al., Ann. Bot. Fenn. 51: 360. 2014. but statistically supported lineages. These incongruences may Type specimens. USA, Oregon, Benton County, Corvallis, on old apple be due to recombination, hybridisation, or incomplete lineage trees, Dec. 1931, F.P. Sipe 669 (BP 33958 – holotype of Alectoria pseudofusce- sorting, as documented in many other species groups (e.g., scens). ― Canada, British Columbia, 25 Sept. 2006, T. Goward 07-02-2011 Jakob & Blattner 2006, McGuire et al. 2007, Edwards et al. (UBC (S222) – epitype selected here, MycoBank MBT381732; British Colum- 2008, Stewart et al. 2014). In lichen-forming fungi, outcrossing bia, Clearwater Valley, 0.5 km S of Philip Creek, ‘Edgewood West’, 715 m, and recombination have been demonstrated, for example, in 9 Nov. 2011, T. Goward 11-61 (UBC (L347) – holotype of Bryoria inactiva). Lobaria pulmonaria (Zoller et al. 1999, Singh et al. 2012, Kel- Nomenclature ― A number of species rank names are syno- ler & Scheidegger 2016), Letharia (Kroken & Taylor 2001a, b, nymised to this taxon, but these have not been epitypified with Altermann et al. 2014), and Cladonia (Steinová et al. 2013). sequenced material. All these names, however, are later in Apothecia are usually absent in Bryoria sect. Implexae, and date than pseudofuscescens, and so could not have priority even when present may not contain mature spores. If cryptic over that name. Further information on type materials can be sexuality is not occurring, hybridization is unlikely to provide an seen in Brodo & Hawksworth (1977) and Velmala et al. (2014). explanation of our data. In the absence of sexual reproduction, Although no samples of Bryoria salazinica have been studied at any recombination is improbable, although some fungi lacking molecular level, the published description and images (Brodo sexual structures show recombination events attributable to & Hawksworth 1977) suggests this taxon also belong here. parasexual cycles (Schoustra et al. 2007). We did detect signals Distribution ― Only known from North America, growing on suggesting putative recombination in the FRBi loci, but not in bark, branches, rock or soil. the standard three loci used in the taxonomy adopted here. Recombination signals may reflect some mitotic recombination, Notes ― Characterised by the absence of soralia and detect- actual or ancient sexual reproduction (Douhan et al. 2007) or be able atranorin. merely false positives produced by chance production of similar sequences. In any case, recombination alone is insufficient DISCUSSION to explain all the discordances found. For example, only one putative recombination event was detected in FRBi15, and Phylogenetic relationships disentangling the FRBi16 recombination points is insufficient Species concepts in Bryoria sect. Implexae have previously to obtain the topology of the three-locus phylogeny. Incongru- been based primarily on well-characterised northern European ences may also be caused by the analysis of different paralogs and North American specimens (Hawksworth 1972, Brodo & of FRBi15 and FRBi16 amplified with the new primers, but this Hawksworth 1977, Velmala et al. 2014). Velmala et al. (2014) seems improbable, as no paralogs have been detected in the recognised 11 species on the basis of morphological and SSRs of these loci (Boluda et al. unpubl.). However, our results chemical characters, but many of these were not supported by indicate recent diversification and large effective population molecular data, and different species names were accepted for sizes in this lichenised complex. Thus, incongruences amongst taxa that could not be distinguished molecularly. We discovered loci seem rather attributable to incomplete lineage sorting. that these demarcations broke down when specimens from The different putative species concepts generated by the spe- more southern European populations were incorporated. This cies delimitation programs (Table 3) may not be only due to is shown in a phenetic analysis using only phenotypically diag- the different algorithms applied, but also because some of the nostic characters (Appendix 2), where the resulting groups are programs were designed for use in single-locus phylogenies not resolved as clear-cut morphospecies. Indeed, any character (e.g., ABGD, PTP, or GMYC). Nevertheless, all the clustering previously used in the group could be used to define the three analyses showed a tendency to distinguish four groups, G, Ko, lineages of the Bryoria fuscescens complex (Fig. 6). NA, and WD (Table 4; Fig. 6). STRUCTURE was unable to C.G. Boluda et al.: Mismatch between phenotypes and genotypes in lichenized fungi 89

Fig. 6 Integrated assessment of results in Bryoria sect. Implexae. Tree topology depicts the result of the Bayesian Markov chain Monte Carlo (B/MCMC) analysis. Posterior probabilities and bootstrap analysis for the supported nodes (≥ 0.95 and ≥ 70 %) are indicated at the main nodes. For each specimen, the extrolites detected, and the putative number of species predicted by the different methodologies is shown. The left top corner tree shows the results of the molecular dating analysis. ― a. Bryoria implexa morphotype (Spain, Asturias, 2013, Boluda, MAF-Lich. 20749); b. B. capillaris morphotype (Spain, Navarra, 2013, Boluda & Villagra, MAF-Lich. 20748); c. B. fuscescens morphotype (Morocco, Ifrane, 2013, Boluda, MAF-Lich. 20751). ― Ale. = Alectorialic acid; Bar. = Barbatolic acid; Fum. = Fumarprotocetraric acid; G = Glabra clade; Gyr. = Gyrophoric acid; Hap. Net. = Haplotype Network; HPD = Highest Posterior Density; Ko = Kockiana clade; Mya = Million years ago; NA = North American clade; Nor. = Norstictic acid; PCoA = Principal Coordinates Analysis; Pso. = Pso- romic acid; STRUCT. = Structure; WD = Wide Distributed clade. 90 Persoonia – Volume 42, 2019 define these groups until reaching the K = 6 hypothesis, which suggest that groups such as WDr, WDg, or pik5 might merit can be attributed to the highly unbalanced sampling sizes; the species rank, but our experience, together with the results from analyses shows that the WD cluster is much more variable other analyses shown here, leads us to reject this hypothesis. than NA, which was not divided into subgroups until K = 10. We conclude that the most pragmatic solution, supported by Specimen 49 (identified as B. pikei, Fig. 2 asterisk), probably the general trend of the results from the different analyses we reflects the impossibility of unequivocally distinguishing that performed is to consider clades G, Ko, NA, and WD as the species from B. capillaris; however, as sequence amplification species Bryoria glabra, B. kockiana, B. pseudofuscescens, of this specimen failed, we cannot determine if this mismatch and B. fuscescens, respectively. was due to misidentification or DNA contamination. Haplotype Clade age can contribute to decisions as to species limits. network analyses have been extensively used in infraspecific Divergence time estimates can be robust if the analyses are population and less frequently closely related species groups performed with well-resolved phylogenies and can incorporate (Houbraken et al. 2012, Pino-Bodas et al. 2016). Even this fossil calibration points, as in some vertebrates (Perelman et type of nested clade phylogeographic analysis has some critics al. 2011). Contrarily, in lichenised fungi, fossils are rare and in (e.g., Knowles 2008, Templeton 2009). The resulting network many cases enigmatic or with ambiguous relationships (Thomas in the present case is concordant with that obtained from other et al. 2014, Hawksworth 2015, Kaasalainen et al. 2015). In ad- analyses. If two DNA barcoding standard marker networks are dition, generation times can be expected to be different among obtained from a single analysis with a 95 % parsimony con- species, as is the case with nuITS locus mutation rates between nection limit, members of each network might be considered herbaceous and woody plants, or even a difference of almost as different species (Hart & Sunday 2007), showing the clear an order of magnitude between different plant genera (Kay isolation of B. glabra from the other taxa in the complex. In the et al. 2006). Here we used a nuITS mutation rate estimated case of connections with several mutation steps, as between from Melanohalea, a genus in the same family (Leavitt et al. clades Ko, NA, and WD, taxon delimitation is below the species 2012a), species of which frequently grow with Bryoria and level, but in any case indicates some degree of genetic isolation. reproduce asexually as well as sexually. The split of B. glabra The relationships amongst the Ko, NA, and WD clades remains from the other taxa in the B. fuscescens complex is estimated unresolved, indicating that the evolutionary history may be too at c. 6.9 Mya, and clearly separated from the much later complex to be adequately captured by dichotomous phylo divergence of the other three species estimated at c. 1 Mya genies based on a few neutral markers. Moreover, the putative (0.6 Mya if only the nuITS locus is used). This contrasts with presence of shared ancestral polymorphisms amongst the other lichenised species considered of recent origin, estimated clades may be producing incompatible topologies, which result around 2.5‒5 Mya (Pliocene), with any Pleistocene speciation in clades with low support. event rare and always older than 1 Mya (Amo de Paz et al. We also performed analyses to estimate changes in past po 2012, Leavitt et al. 2012a, b, Molina et al. 2016). As the three pulation sizes, which may have affected current clade diver- B. fuscescens complex clades seem to have diverged more sity. Genealogies of most plant and animal species coalesce recently, the extent of their reproductive isolation is uncertain, between 2.58‒0.01 Mya (Grant 2015), and our estimated and the discovery of intermediate lineages amongst other intervals are within this range. However, in our case the dates named species from unsampled geographical regions, such are relatively recent, with the oldest coalescences at 0.7 Mya. as continental Asia remains possible. A flat graphic is generally interpreted as population stability In the absence of supporting phenotypic, geographic, or eco- but can also be due to a lack of detection power produced by logical differences, the recent divergence, and the possibility small sample sizes. Moreover, a small rise in the curve near the of incomplete lineage sorting, clades Ko, NA, and WD may be present, seen in Fig. 5, can be a consequence of the random considered as conspecific evolving lineages. It is, however, sampling of the MCMC haplotype trees (Grant 2015); this result important to recognise the lineages formally in order to facilitate must therefore be treated with caution. Our sequences do not their conservation by enabling their threat status to be assessed bear imprints of ancient population history, but rather more re- by IUCN criteria. We decided not to adopt the rank of subspe- cent population growth, for example by extensions northwards cies as that is now hardly used in mycology, and then not in in post-glacial periods. The loss of information may also arise any consistent way; traditionally this rank was used in plants for from bottlenecks (i.e., a marked reduction in population size), morphologically distinguishable populations with geographical local extinctions, and subsequent recolonization. Additionally, differences and where intermediates occurred where they were the use of genes with low levels of polymorphisms, may impede sympatric (Stuessy 2009). a robust reconstruction of population sizes through time. The formal recognition of cryptic lineages at species level, as Species concept suggested by our analyses, emerges as the most appropriate solution. Cryptic speciation is now recognised as a common Some species delimitation analyses, such as STRUCTURE, phenomenon in Parmeliaceae, and our results are in accord GMYC, PTP, can overestimate the number of taxa meriting with other studies in which molecular markers in combination formal recognition, particularly when sampling is uneven or in with statistical tools revealed genetically distinct lineages pre- species with a strong intraspecific genetic structure (Altermann viously hidden under a single taxon name in this family (e.g., et al. 2014, Modica et al. 2014, Alors et al. 2016, Del-Prado et al. 2016). ABGD, in contrast, has been considered as rather Singh et al. 2015, Alors et al. 2016, Del-Prado et al. 2016, conservative, less prone to species overestimation and less Divakar et al. 2016, Leavitt et al. 2016). Further, this solution sensitive to unbalanced sampling. While that method only de- is in line with the increasing need to formally recognize cryp- tects discontinuities in DNA sequence variation (Puillandre et tic species-level lineages in all fungi (Hibbett 2016); indeed, al. 2011), it can also be expected to fail in species with strong cryptic speciation may mean that there are on average ten or population genetics structures, for example ones containing more fungi masked in formally named species (Hawksworth & exclusive haplotypes. All species delimitation programs will Lücking 2017). provide a number of reasonably discrete groups (‘evolving line- Of the lineages recognised here, only the WD clade emerged ages’) that should be evaluated for consideration as meriting as cosmopolitan, occurring in Europe, Asia, North America, and species rank, but the decision has to be by taxonomists with Africa (Appendix 5). NA and Ko have been collected so far only experience in the group concerned. Some of our analyses in North America, despite our extensive sampling in Europe C.G. Boluda et al.: Mismatch between phenotypes and genotypes in lichenized fungi 91

(Appendix 5). Further sampling, especially in South America, Material referred to the ‘capillaris’ and ‘fuscescens’ pheno- Asia, and Africa, is needed before any finer-scale biogeographic types has been reported to show slight differences in water patterns might be detected. holding capacity (Esseen et al. 2015), and also the pigments The practical issue of naming older museum specimens and may provide protection against excesses of light (Färber et al. material in ecological surveys could be resolved by recognis- 2014). Further, in southern Europe particularly, the ‘capillaris’ ing the three groups as species within a broad concept, such phenotype tends to be restricted to humid, shaded, and pro- as an aggregate, complex, or adding ‘s.lat.’. We considered tected or undisturbed environments than the ‘fuscescens’ one, commending the adoption of the suffix ‘agg.’ for material when something already recognised by Motyka (1964). Additionally, precise molecular species identifications cannot be made. While in northern Europe, dark specimens containing barbatolic acid this has been done in a few other groups of fungi (e.g., Parn- are much more common than in southern Europe, where they men et al. 2013, Pažoutová et al. 2015), ‘complex’ has come to are extremely rare (Myllys et al. 2016, Esseen et al. 2017). As be used more widely and was strongly favoured at the Cryptic both phenotypes can grow side by side and even intermixed Speciation in Fungi symposium in Utrecht in September 2017 on the same trees, where environmental conditions must be (report awaited). We therefore suggest the use of ‘complex’ here identical, ecological plasticity has to be discounted. However, but recognise some may prefer to use ‘agg.’ or ‘s.lat.’. Where some unknown epigenetic modification could perhaps have a DNA samples can be obtained and analysed, we recommend role in that process, as once a metabolic pathway is activated use of the GAPDH locus, as all the other tested markers are or silenced, it may be hardly modifiable under more or less not able to distinguish with confidence the three species we neutral environmental conditions, transferring the phenotypes recognize in the complex. to the clonal offspring. Specimens with dark thalli, containing barbatolic acid, or with pale thalli with traces of barbatolic and Infraspecific phenotypic variation also containing other extrolites, could represent transitional specimens. While our results support rejection of the morphospecies concept in this group of lichens, two main phenotypes can nev- Molecular and morphological rates of divergence may some- ertheless often be distinguished by the naked eye in the field: times be uncoupled. Incomplete lineage sorting arises when 1. the pale grey ‘capillaris’ morphotype (including B. capillaris an ancestral polymorphism persists through a speciation event and B. pikei, Fig. 6b); and and each polymorphism can lead to different alleles being car- 2. the fuscous brown to dark brown ‘fuscescens’ morphotype, ried among descendants (Maddison 1997, Hartl & Clark 2007). in which most other species names are placed (Fig. 6a, c). Consequently, different tree topologies may be obtained depending which specimens or loci are used. Rosenberg (2003) The chemical characters are not always checked by field has shown that 5.3N generations are needed for a species to workers, and while the ‘capillaris’ morphotype typically has e acquire monophyly at 99 % of its loci given that all loci in the benzyldepsides, the ‘fuscescens’ morphotype lacks those sister species are also monophyletic. That indicates that for a compounds and has fumarprotocetraric acid or various depsi- species of 1 M individuals with a generation time of 10 yr, the dones. However, there are dark morphs with benzyldepsides full monophyly will only be reached 50 M years after speciation, (once called f. fuscidula), and pale grey ones with fumarpro- whereas only around 1 000 yr may be needed for species with tocetraric acid (e.g., B. subcana) or other depsidones (e.g., B. small populations (Naciri & Linder 2015). Incomplete lineage kuemmerleana). It is conceivable that the two morphotypes sorting may be frequent in closely related taxa or during a spe- originated before the separation of B. pseudofuscescens and ciation process (Hobolth et al. 2011, Blanco-Pastor et al. 2012, B. fuscescens, as both colour variants and chemistries appear Saag et al. 2014, Naciri & Linder 2015), as may be considered in both taxa. This phenomenon cannot be explained by a simple in our case. The topological incongruence observed among the ongoing speciation event in which one lineage has originated standard loci, FRBi15 and FRBi16, supports the incomplete line- new adaptations, but is still not isolated from the parental age sorting hypothesis as one of the main reasons explaining lineages, as neither are monophyletic in a paraphyletic clade. why morphospecies are not monophyletic. While neutral mark- The difference in phenotype cannot be attributed to different ers are useful for understanding gene-flow patterns, adaptive algal partners as all material in the complex shares the same markers provide the evolutionary pressure that contributes to species and even in many cases the same nuITS haplotypes of speciation (Emelianov et al. 2004, Hey 2006, Holderegger et al. Trebouxia (Lindgren et al. 2014, Boluda et al. unpubl.). Further, 2006). As adaptive markers are under natural selection, certain as we used neutral markers to detect gene-flow gaps between alleles can be present in some morphospecies and absent in lineages, the phenotypes are also not the result of genetic isola- others, even if there is gene flow amongst them (Lumley & tion, and other possibilities must be considered. Sperling 2011). The use of phylogenomic datasets may provide It has recently been reported that yeast morphs of the lichenicol- a more accurate and supported phylogenetic reconstruction, ous and gall-forming basidiomycete genus Cyphobasidium can especially if the appropriate scale of loci variability is selected be abundant in or on the outermost cortical tissues of Bryoria from all the genome (Leavitt et al. 2016). However, if there species (Spribille et al. 2016). Spribille et al. (2016) reported are high levels of incomplete lineage sorting, it might not be a possible relation between Cyphobasidium yeast abundance expected that morphospecies would appear forming supported and vulpinic acid production in two other species of Bryoria, clades. Nevertheless, genomic data may reveal few mutations B. fremontii and B. tortuosa, and also visualised these yeasts linked to certain morphospecies, which would be producing in material identified as B. capillaris phenotypes. Contrary to adaptive traits. Darwin’s finches are an iconic example of a rapid the claims of Spribille et al. (2016), these fungi do not appear speciation process, in which there is a mismatch between the to be an integral part of the mutualism (Oberwinkler 2017). It phylogenetic species concept and phenotype-based taxonomy; is, however, feasible that the yeasts cells are able to develop in that case, genomic studies have detected specific loci sub- to a greater extent in ‘capillaris’ morphotypes as the cortices jected to selection pressure, which are directly related with the can have lumpier polysaccharide deposits than do those of development of taxon-specific phenotypes (Lamichhaney et al. ‘fuscescens’ (Hawksworth 1969b, Boluda et al. 2014, Esseen 2015). In Bryoria, supposed adaptive traits may be influenced et al. 2017). How the occurrence of yeast morphs of these by the genes involved in the production of certain extrolites or lichenicolous fungi in the surface of the cortex could possibly in the epicortical substances (Boluda et al. 2014), which may determine colour morphotypes is obscure. produce differential selection pressure for each morphotype, 92 Persoonia – Volume 42, 2019 at least in some environments. The process might be similar Dal Grande F, Alors D, Pradeep KD, et al. 2014. Insights into intrathalline to that of natural selection of the pale and melanic morphs genetic diversity of the cosmopolitan lichen symbiotic green alga Trebouxia of the Peppered Moth (Biston betularia) in Europe (Majerus decolorans Ahmadjian using microsatellite markers. Molecular Phylogene 2009), impeding the fixation of a single morphotype in all tics and Evolution 72: 54‒60. Darriba D, Taboada GL, Doallo R, et al. 2012. jModelTest 2: more models, populations. In our case also, high levels of incomplete lineage new heuristics and parallel computing. Nature Methods 9: 772. sorting mixed with a few phenotypically important genes under Dayrat B. 2005. Towards integrative taxonomy. Biological Journal of the variable degrees of selection in different environments, may Linnean Society 85: 407‒415. explain the mismatch observed between phenotypes and geno- Del-Prado R, Divakar PK, Lumbsch HT, et al. 2016. Hidden genetic diversity types. in an asexually reproducing lichen forming fungal group. PLoS ONE 11: e0161031. Divakar PK, Cubas P, Blanco O, et al. 2010. An overview on hidden diversity Acknowledgements This work was undertaken with the support of the Span- in lichens: Parmeliaceae. http://taxateca.com/files/Divakar_et_al_2010.pdf. ish Ministerio de Economía y Competitividad projects CGL2011-25003 and Divakar PK, Leavitt SD, Molina MC, et al. 2016. A DNA barcoding approach CGL2014-55542-P, and the BES-2012-054488 grant to CGB. Microsatellite for identification of hidden diversity in Parmeliaceae (Ascomycota): Parmelia analyses carried out at WSL were financially supported by the Federal Office sensu stricto as a case study. Botanical Journal of the Linnean Society for the Environment (FOEN) and SwissBOL (grants to CS), and we acknow 180: 21‒29. ledge the Genetic Diversity Centre at ETHZ. We are grateful also to A. Crespo Douhan GW, Martin DP, Rizzo DM. 2007. Using the putative asexual fungus (Madrid), B. Abbott (Arkadias), C. Ruibal (Madrid), H. Holien (Steinkjer), Cenococcum geophilum as a model to test how species concepts influence J. Villagra (Madrid), M. Wedin (Stockholm), N. Calpena-Grau (Madrid), and recombination analyses using sequence data from multiple loci. Current T. Goward (Vancouver) for providing specimens. We are especially grateful Genetics 52: 191‒201. to C. Cornejo (Switzerland) for her help with the laboratory work, and for a Drumond AJ, Rambaut A, Shapiro B, et al. 2005. Bayesian coalescent infer- helpful review of the manuscript. ence of past population dynamics from molecular sequences. Molecular Biology and Evolution 22: 1185‒1192. REFERENCES Drumond AJ, Suchard MA, Xie D, et al. 2012. Bayesian phylogenetics with BEAUti and the BEAST 1.7. Molecular Biology and Evolution 29: 1969‒ Abbott RJ, Barton NH, Good JM. 2016. Genomics of hybridization and its 1973. evolutionary consequences. Molecular Ecology 25: 2325‒2332. Earl DA, Von Holdt B. 2012. STRUCTURE HARVESTER: a website and Akaike H. 1974. A new look at the statistical model identification. IEEE Trans program for visualizing STRUCTURE output and implementing the Evanno actions on Automatic Control 19: 716‒723. method. Conservation of Genetics Resources 4: 359‒361. Alors D, Lumbsch TH, Divakar PK, et al. 2016. An integrative approach for Edwards CE, Soltis DE, Soltis PS. 2008. Using patterns of genetic structure understanding diversity in the Punctelia rudecta species complex (Parme- based on microsatellite loci to test hypotheses of current hybridization, liaceae, Ascomycota). PLoS ONE 11: e0146537. ancient hybridization and incomplete lineage sorting in Conradina (Lami- Altermann S, Leavitt SD, Goward T, et al. 2014. How do you solve a problem caceae). Molecular Ecology 17: 5157‒5174. like Letharia? A new look at cryptic species in lichen-forming fungi using Emelianov I, Marec F, Mallet J. 2004. Genomic evidence for divergence with Bayesian clustering and SNPs from multilocus sequence data. PLoS ONE gene flow in host races of the larch budmoth. Proceedings of the Royal 5: e97556. Society of London, Biological Sciences 271: 97‒105. Amo de Paz G, Cubas P, Crespo A, et al. 2012. Transoceanic dispersal and Esseen PA, Olsson T, Coxson D, et al. 2015. Morphology influences water subsequent diversification on separate continents shaped diversity of the storage in hair lichens from boreal forest canopies. Fungal Ecology 18: Xanthoparmelia pulla group (Ascomycota). PLoS ONE 7: e39683. 26‒35. Articus K, Mattsson J-E, Tibell L, et al. 2002. Ribosomal DNA and β-tubulin Esseen PA, Rönnqvist M, Gauslaa Y, et al. 2017. Externally held water – a key data do not support the separation of the lichens Usnea florida and factor for hair lichens in boreal forest canopies. Fungal Ecology 30: 29‒38. U. subfloridana as distinct species. Mycological Research 106: 412‒418. Evanno G, Regnaut S, Goudet J. 2005. Detecting the number of clusters of Bekessy SA, Ennos RA, Burgman MA, et al. 2003. Neutral DNA markers fail individuals using the software STRUCTURE: a simulation study. Molecular to detect genetic divergence in an ecologically important trait. Biological Ecology 14: 2611‒2620. Conservation 110: 267‒275. Falush D, Stephens M, Pritchard J. 2003. Inference of population structure Blanco-Pastor JL, Vargas P, Pfeil BE. 2012. Coalescent simulations reveal using multilocus genotype data: Linked loci and correlated allele frequen- hybridization and incomplete lineage sorting in Mediterranean Linaria. cies. Genetics 155: 945‒959. PLoS ONE 7: e39089. Färber L, Solhaug KA, Esseen PA, et al. 2014. Sunscreening fungal pig- Boluda CG, Hawksworth DL, Divakar PK, et al. 2016. Microchemical and ments influence the vertical gradient of pendulous lichens in boreal forest molecular investigations reveal Pseudephebe species as cryptic with an canopies. Ecology 95: 1464‒1471. environmentally modified morphology. The Lichenologist 48: 527‒543. Fu YX. 1997. Statistical test of neutrality of mutations against population Boluda CG, Rico VJ, Crespo A, et al. 2015. Molecular sequence data from growth, hitchhiking and background selection. Genetics 147: 915‒925. populations of Bryoria fuscescens s.lat in the mountains of central Spain Gardes M, Bruns TD. 1993. ITS primers with enhanced specificity for indicates a mismatch between haplotypes and chemotypes. The Lichenolo- basidiomycetes-application to the identification of mycorrhizae and rusts. gist 47: 279‒286. Molecular Ecology 2: 113–118. Boluda CG, Rico VJ, Hawksworth DL. 2014. Fluorescence microscopy as Gargas A, Taylor JW. 1992. Polymerase chain reaction (PCR) primers for a tool for the visualization of lichen substances within Bryoria thalli. The amplifying andsequencing nuclear 18s rDNA from lichenized fungi. Myco- Lichenologist 46: 723‒726. logia 84: 589–592. Boni MF, Posada D, Feldman MW. 2007. An exact nonparametric method for Gibbs MJ, Armstrong JS, Gibbs AJ. 2000. Sister-scanning: a Monte Carlo inferring mosaic structure in sequence triplets. Genetics 176: 1035‒1047. procedure for assessing signals in recombinant sequences. Bioinforma Brodo IM, Hawksworth DL. 1977. Alectoria and allied genera in North America. tics 16: 573‒582. Opera Botanica 42: 1‒164. Grant WS. 2015. Problems and cautions with sequence mismatch analysis Campbell V, Legendre P, Lapointe FJ. 2011. The performance of the congru- and Bayesian skyline plots to infer historical demography. Journal of Here- ence among distance matrices (CADM) test in phylogenetic analysis. BMC dity 16: 1‒14. Evolutionary Biology 11: 64. Hart WM, Sunday J. 2007. Things fall apart: biological species form uncon- Caparrós R, Lara F, Draper I, et al. 2016. Integrative taxonomy sheds light nected parsimony networks. Biology Letters 3: 509‒512. on an old problem: the Ulota crispa complex (Orthotrichaceae, Musci). Hartl DL, Clark AG. 2007. Principles of population genetics. 4th Ed. Sinauer Botanical Journal of the Linnean Society 180: 427‒451. Associates Inc, Sunderland, MA. Chatrou LW, Escribano MP, Viruel MA, et al. 2009. Flanking regions of mono- Hawksworth DL. 1969a. Chemical and nomenclatural notes on Alectoria morphic microsatellite loci provide a new source of data for plant species- (Lichenes) I. Taxon 18: 393–399. level phylogenetics. Molecular Phylogenetics and Evolution 53: 726‒733. Hawksworth DL. 1969b. The scanning electron microscope, an aid to the Clement M, Posada D, Crandall KA. 2000. TCS: a computer program to study of cortical hyphal orientation in the lichen genera Alectoria and estimate gene genealogies. Molecular Ecology 9: 1657‒1659. Cornicularia. Journal de Microscopie 8: 753‒760. Crespo A, Lumbsch HT. 2010. Cryptic species in lichen-forming fungi. IMA Hawksworth DL. 1972. Regional studies in Alectoria (Lichenes) II. The British Fungus 1: 167‒170. species. The Lichenologist 5: 181‒261. C.G. Boluda et al.: Mismatch between phenotypes and genotypes in lichenized fungi 93

Hawksworth DL. 1976. Lichen chemotaxonomy. In: Brown DH, Hawksworth Lanfear R, Calcott B, Ho SY, et al. 2012. Partitionfinder: combined selection DL, Bailey LH (eds), Lichenology: progress and problems: 139‒184. Aca of partitioning schemes and substitution models for phylogenetic analyses. demic Press, London. Molecular Biology and Evolution 29: 1695‒1701. Hawksworth DL. 2015. Lichenization: The origins of fungal life-style. In: Leavitt SD, Divakar PK, Crespo A, et al. 2016. A matter of time ‒ under- Upreti DK, Divakar PK, Shukla V, et al (eds), Recent advances in licheno standing the limits of the power of molecular data for delimiting species logy. Modern methods and approaches in lichen systematics and culture boundaries. Herzogia 29: 479‒492. techniques, vol. 2: 1‒10. Springer, India. Leavitt SD, Esslinger TL, Divakar PK, et al. 2012a. Miocene and Pliocene Hawksworth DL, Jørgensen PM. 2013. (2196) Proposal to conserve the dominated diversification of the lichen-forming fungal genus Melanohalea name Alectoria fuscescens (Bryoria fuscescens) against Lichen chalybei- (Parmeliaceae, Ascomycota) and Pleistocene population expansions. BMC formis and Alectoria subcana (Ascomycota: Lecanorales: Parmeliaceae). Evolutionary Biology 12: 176. Taxon 62: 1057. Leavitt SD, Esslinger TL, Lumbsch HT. 2012b. Neogene-dominated diversi Hawksworth DL, Lücking R. 2017. Fungal diversity revisited: 2.2 to 3.8 million fication in neotropical montane lichens: Dating divergence events in the species. Microbiology Spectrum 5: FUNK-0052-2016. lichen-forming fungal genus Oropogon (Parmeliaceae). American Journal Hey J. 2006. Recent advances in assessing gene flow between diverging of Botany 99: 1764‒1777. populations and species. Current Opinion in Genetics & Development Leavitt SD, Grewe F, Widhelm T, et al. 2016. Resolving evolutionary rela- 16: 592‒596. tionships in lichen-forming fungi using diverse phylogenomic datasets and Hibbett DS. 2016. The invisible dimension of fungal diversity. Science 351: analytical approaches. Scientific Reports 6: 22262. 1150‒1151. Leavitt SD, Johnson L, St. Clair LL. 2011. Species delimitation and evolution Hobolth A, Dutheil JY, Hawks J, et al. 2011. Incomplete lineage sorting pat- in morphologically and chemically diverse communities of the lichen-forming terns among human, chimpanzee, and orangutan suggest recent orangutan genus Xanthoparmelia (Parmeliaceae, Ascomycota) in Western North speciation and widespread selection. Genome Research 21: 349‒356. America. American Journal of Botany 98: 175‒188. Holderegger R, Kamm U, Gugerli F. 2006. Adaptative vs neutral genetic diver- Legendre P, Lapointe FJ. 2004. Assessing congruence among distance sity: implications for landscape genetics. Landscape Ecology 21: 797‒807. matrices: Single-malt Scotch whiskies revisited. Australian & New Zealand Houbraken J, Frisvad JC, Seifert KA, et al. 2012. New penicillin-producing Journal of Statistics 46: 615‒629. Penicillium species and an overview of section Chrysogena. Persoonia Librado P, Rozas J. 2009. DnaSP v5: A software for comprehensive analysis 29: 78‒100. of DNA polymorphism data. Bioinformatics 25: 1451‒1452. Huelsenbeck JP, Ronquist F. 2001. MrBayes: Bayesian inference of phylo- Lindgren H, Velmala S, Högnabba F, et al. 2014. High fungal selectivity for genetic trees. Bioinformatics 17: 754‒755. algal symbionts in the genus Bryoria. The Lichenologist 46: 681‒695. Jakob SS, Blattner FR. 2006. A chloroplast genealogy of Hordeum (Poaceae): Lumbsch HT. 1988. The use of metabolic data in lichenology at the species Long-term persisting haplotypes, incomplete lineage sorting, regional and subspecific levels. The Lichenologist 30: 357‒367. extinction, and the consequences for phylogenetic inference. Molecular Lumley LM, Sperling FAH. 2011. Utility of microsatellites and mitochondrial Biology and Evolution 23: 1602‒1612. DNA for species delimitation in the spruce budworm (Choristoneura fumi Jakobsson M, Rosenberg NA. 2007. CLUMMP: a cluster matching and ferana) species complex (Lepidoptera: Tortricidae). Molecular Phylogenet- permutation program for dealing with label switching and multimodality in ics and Evolution 58: 232‒243. analysis of population structure. Bioinformatics 23: 1801‒1806. Lutsak T, Fernández-Mendoza F, Nadyeina O, et al. 2017. Testing the correla- Joly S, McLenachan PA, Lockhart PJ. 2009. A statistical approach for dis- tion between norstictic acid content and species evolution in the Cetraria tinguishing hybridization and incomplete lineage sorting. The American aculeata group in Europe. The Lichenologist 49: 39‒56. Naturalist 174: E54‒E70. Maddison WP. 1997. Gene trees in species trees. Systematic Biology 46: Jones G, Aydin Z, Oxelman B. 2014. DISSECT: an assignment-free Bayesian 523‒536. discovery method for species delimitation under the multispecies coales- Maechler M, Rousseeuw P, Struyf A, et al. 2013. cluster: Cluster Analysis cent. Bioinformatics 31: 991‒998. Basics and Extensions. R package version 1.14.4. Jørgensen PM, Galloway DJ. 1983. Bryoria (lichenized Ascomycota) in New Majerus MEN. 2009. Industrial melanism in the Peppered Moth, Biston Zealand. New Zealand Journal of Botany 21: 335‒340. betularia: An excellent teaching example of Darwinian evolution in action. Kaasalainen U, Heinrichs J, Krings M, et al. 2015. Alectorioid morphologies Evolution: Education and Outreach 2: 63‒74. in Paleogene lichens: new evidence and re-evaluation of the fossil Alectoria Mark K, Saag L, Leavitt SD, et al. 2016. Evaluation of traditionally circum- succini Mägdefrau. PLoS ONE 10: e0129526. scribed species in the lichen-forming genus Usnea, section Usnea (Par- Katoh K, Standley DM. 2013. MAFFT multiple sequence alignment software meliaceae, Ascomycota) using six-locus dataset. Organisms Diversity and version 7: improvements in performance and usability. Molecular Biology Evolution 16: 497‒524. and Evolution 30: 772‒780. Martin DP, Lemey P, Lott M, et al. 2010. RDP3: a flexible and fast computer Kay KM, Whittall JB, Scott AH. 2006. A survey of nuclear ribosomal internal program for analyzing recombination. Bioinformatics 26: 2462‒2463. transcribed spacer substitution rates across angiosperms: an approximate Martin DP, Posada D, Crandall KA, et al. 2005. A modified bootscan algorithm molecular clock with life history effects. BMC Evolution Biology 6: 36. for automated identification of recombinant sequences and recombination Keller C, Scheidegger C. 2016. Multiple mating events and spermatia-medi breakpoints. AIDS Research and Human Retroviruses 21: 98‒102. ated gene flow in the lichen-forming fungus Lobaria pulmonaria. Herzogia Martin DP, Rybicki E. 2000. RDP: detection of recombination amongst aligned 29: 435–450. sequences. Bioinformatics 16: 562‒563. Keller NP, Hohn TM. 1997. Metabolic pathway gene clusters in filamentous Maynard-Smith J. 1992. Analyzing the mosaic structure of genes. Journal fungi. Fungal Genetics and Biology 21: 17‒21. of Molecular Evolution 34: 126‒129. Kirika P, Divakar PK, Crespo A, et al. 2016a. Polyphyly of the genus Cano- McGuire JA, Linkem CW, Koo MS, et al. 2007. Mitochondrial introgression parmelia ‒ uncovering incongruences between phenotype-based classifi and incomplete lineage sorting through space and time: Phylogenetics of cation and molecular phylogeny within lichenized Ascomycota (Parmelia crotaphytid lizards. Evolution 61: 2879‒2897. ceae). Phytotaxa 289: 36‒48. McMullin RT, Lendemer JC, Braid HE, et al. 2016. Molecular insights into the Kirika P, Divakar PK, Crespo A, et al. 2016b. Phylogenetic studies uncover a lichen genus Alectoria (Parmeliaceae) in North America. Botany 94: 1‒11. predominantly African lineage in a widely distributed lichen-forming fungal Miller MA, Pfeiffer W, Schwartz T. 2010. Creating the CIPRES Science species. MycoKeys 14: 1‒16. Gateway for inference of large phylogenetic trees. In: Proceedings of the Knowles LL. 2008. Why does a method that fails continue to be used. Evolu- Gateway Computing Environments Workshop (GCE), November 14: 1‒8. tion 62: 2713‒2717. USA, New Orleans. Konrad H, Kiristis T, Riegler M, et al. 2002. Genetic evidence for natural Modica MV, Puillandre N, Castelin M, et al. 2014. A good compromise: rapid hybridization between the Dutch elm disease pathogens Ophiostoma and robust species proxies for inventorying biodiversity hotspots using the novo-ulmi ssp. novo-ulmi and O. novo-ulmi ssp. americana. Plant Patho Terebridae (Gastropoda: Conoidea). PLoS ONE: e102160. logy 51: 78‒84. Molina MC, Del-Prado R, Divakar PK, et al. 2011a. Another example of cryptic Kroken S, Taylor JW. 2001a. Outcrossing and recombination in the lichenized diversity in lichen-forming fungi: the new species Parmelia mayi (Ascomy- fungus Letharia. Fungal Genetics and Biology 34: 83–92. cota: Parmeliaceae). The Lichenologist 11: 331–342. Kroken S, Taylor JW. 2001b. A gene genealogical approach to recognize phylo- Molina MC, Divakar PK, Goward T, et al. 2016. Neogene diversification in genetic species boundaries in the lichenized fungus Letharia. Mycologia the temperate lichen-forming fungal genus Parmelia (Parmeliaceae, Asco 93: 38–53. mycota). Systematics and Biodiversity 15: 166‒181. Lamichhaney S, Berglund J, Markus SA, et al. 2015. Evolution of Darwin’s Molina MC, Divakar PK, Millanes AM, et al. 2011b. Parmelia sulcata (Asco- finches and their beaks revealed by genome sequencing. Nature 518: mycota: Parmeliaceae), a sympatric monophyletic species complex. The 371‒375. Lichenologist 43: 585–601. 94 Persoonia – Volume 42, 2019

Monaghan MT, Wild R, Elliot M, et al. 2009. Accelerated species inventory Singh G, Dal Grande F, Cornejo C, et al. 2012. Genetic basis of self-incom on Madagascar using coalescent-based models of species delineation. patibility in the lichen-forming fungus Lobaria pulmonaria and skewed Systematic Biology 58: 298‒311. frequency distribution of mating-type idiomorphs: Implications for conser- Motyka J. 1964. The North American species of Alectoria. The Bryologist vation. PloS ONE 7: e51402. 67: 1‒44. Singh G, Dal Grande F, Divakar PK, et al. 2015. Coalescent-based species Myllys L, Lindgren H, Aikio S, et al. 2016. Chemical diversity and ecology delimitation approach uncovers high cryptic diversity in the cosmopolitan of the genus Bryoria section Implexae (Parmeliaceae) in Finland. The lichen-forming fungal genus Protoparmelia (Lecanorales, Ascomycota). Bryologist 119: 29‒38. PLoS ONE 10: e0124625. Myllys L, Stenroos S, Thell A. 2002. New genes for phylogenetic studies of Spribille T, Touvinen V, Resl P, et al. 2016. Basidiomycete yeasts in the lichenized fungi: glyceraldehyde-3-phosphate dehydrogenase and beta- cortex of ascomycete macrolichens. Science. doi: https://doi.org/10.1126/ tubulin genes. The Lichenologist 34: 237–246. science.aaf8287. Myllys L, Velmala S, Holien H, et al. 2011. Phylogeny of the genus Bryoria. Stamatakis A. 2006. RaxML-VI-HPC: maximum likelihood-based phyloge- The Lichenologist 6: 617‒638. netic analyses with thousands of taxa and mixed models. Bioinformatics Naciri Y, Linder HP. 2015. Species delimitation and relationships: The dance 22: 2688‒2690. of seven veils. Taxon 64: 3‒16. Stamatakis A. 2014. RaxML version 8: A tool for phylogenetic analysis and Nadyeina O, Cornejo C, Boluda CG, et al. 2014. Characterization of micro post-analysis of large phylogenies. Bioinformatics 30: 1312‒1313. satellite loci in lichen-forming fungi of Bryoria section Implexae (Parmeli- Stamatakis A, Hoover P, Rougemont J. 2008. A rapid bootstrap algorithm for aceae). Applied Plant Sciences 2: 1400037. the RAxML webservers. Systematic Biology 57: 758‒771. Nichols R. 2001. Gene trees and species trees are not the same. Trends in Steinová J, Stenroos S, Grube M, et al. 2013. Genetic diversity and spe- Ecology & Evolution 16: 358‒364. cies delimitation of the zeorin-containing red-fruited Cladonia species Oberwinkler F. 2017. Yeasts in Pucciniomycotina. Mycological Progress (lichenized Ascomycota) assessed with ITS rDNA and β-tubulin data. The 16: 831‒856. Lichenologist 45: 665‒684. Orange A, James PW, White FJ. 2010. Microchemical methods for the identi Stewart JE, Timmer LW, Lawrence CB, et al. 2014. Discord between morpho- fication of lichens. 2 edn. British Lichen Society, London. logical and phylogenetic species boundaries: incomplete lineage sorting Padidam M, Sawyer S, Fauquet CM. 1999. Possible emergence of new and recombination results in fuzzy species boundaries in an asexual fungal geminiviruses by frequent recombination. Virology 265: 218‒225. pathogen. BMC Evolutionary Biology 14: 38. Paradis E, Claude J, Strimmer K. 2004. APE: analyses of phylogenetics and Stuessy TF. 2009. Plant taxonomy. The systematic evaluation of comparative data. Columbia University Press, New York. evolution in R language. Bioinformatics 20: 289‒290. Suzuki R, Shimodaira H. 2006. pvclust: an R package for assessing the Parnmen S, Leavitt SD, Rangsiruji A, et al. 2013. Identification of species in uncertainty in hierarchical clustering. Bioinformatics 22: 1540‒1542. the Cladia aggregata group using DNA barcoding (Ascomycota: Lecano- Tajima F. 1989. Statistical method for testing the neutral mutation hypothesis rales). Phytotaxa 115: 1‒14. by DNA polymorphism. Genetics 123: 585‒595. Pažoutová S, Pešicová K, Chudíčková M, et al. 2015. Delimitation of cryptic Tamura K, Peterson D, Peterson N, et al. 2011. MEGA5: Molecular evolu- species inside Claviceps purpurea. Fungal Biology 119: 7‒26. tionary genetics analysis using maximum likelihood, evolutionary distance, Perelman P, Johnson WE, Roos C, et al. 2011. A molecular phylogeny of and maximum parsimony methods. Molecular Biology and Evolution 28: living primates. PLoS Genetics 7: e1001342. 2731‒2739. Pino-Bodas R, Pérez‒Vargas I, Stenroos S, et al. 2016. Sharpening the Templeton AR. 2009. Why does a method that fails continue to be used: the species boundaries in the Cladonia mediterranea complex (Cladoniaceae, answer. Evolution 63: 807‒812. Ascomycota). Persoonia 37: 1‒12. Thomas NT, Krings M, Taylor EL. 2014. Fossil fungi. Academic Press, London. Pons J, Barraclough TG, Gómez-Zurita J, et al. 2006. Sequence based spe- Turland NJ, Wiersema JH, Barrie FR, et al. (eds). 2018. International Code cies delimitation for the DNA taxonomy of undescribed insects. Systematic of Nomenclature for algae, fungi, and plants (Shenzhen Code) adopted Biology 55: 595‒609. by the Nineteenth International Botanical Congress Shenzhen, China, July Posada D, Crandall KA. 2001. Evaluation of methods for detecting recom- 2017. [Regnum Vegetabile no. 159.] Königstein, Koeltz Botanical Books. bination from DNA sequences: computer simulations. Proceedings of Vanhaecke D, García de Leaniz C, Gajardo G, et al. 2012. DNA barcoding the National Academy of Sciences of the United States of America 98: and microsatellites help species delimitation and hybrid identification in 13757‒13762. endangered galaxiid fishes. PLoS ONE 7: e32939. Printzen C, Ekman S. 2002. Genetic variability and its geographical distri- Velmala S, Myllys L, Goward T, et al. 2014. Taxonomy of Bryoria section bution in the widely disjunct Cavernularia hultenii. The Lichenologist 34: Implexae (Parmeliaceae, Lecanoromycetes) in North America and Europe, 101–111. based on chemical, morphological and molecular data. Annales Botanici Pritchard JK, Stephens M, Donnelly P. 2000. Inference of population structure Fennici 51: 345‒371. using multilocus genotype data. Genetics 155: 945‒959. Weiller GF. 1998. Phylogenetic profiles: a graphical method for detecting Puillandre N, Lambert A, Brouillet S, et al. 2011. ABGD, Automatic Barcode genetic recombination in homologous sequences. Molecular Biology and Gap Discovery for primary species delimitation. Molecular Ecology 21: Evolution 15: 326‒335. 1864‒1877. White TJ, Bruns T, Lee S, et al. 1990. Amplification and direct sequencing R Core Team. 2014. R: A language and environment for statistical computing. of fungal ribosomal RNA genes for phylogenetics. In: Innis MA, Gelfand Vienna, R Foundation for Statistical Computing. DH, Sninsky JJ, et al. (eds), PCR protocols: a guide to methods and ap- Rambaut A. 2009. FigTree v.1.4. http://tree.bio.ed.ac.uk/software/figtree/. plications: 315–322. New York: Academic Press. Rambaut A, Suchard MA, Xie D, et al. 2014. Tracer. Version 1.6. http://beast. Widmer I, Dal Grande F, Excoffier L, et al. 2012. European phylogeography of bio.ed.ac.uk/Tracer. the epiphytic lichen fungus Lobaria pulmonaria and its green algal symbiont. Ronquist F, Huelsenbeck JP. 2003. MrBayes 3: Bayesian phylogenetic infer- Molecular Ecology 21: 5827‒5844. ence under mixed models. Bioinformatics 19: 1572‒1574. Will KW, Mishler BD, Wheeler QD. 2005. The perils of DNA barcoding and Rosenberg NA. 2003. The shapes of neutral gene genealogies in two spe- the need for integrative taxonomy. Systematic Biology 54: 844‒851. cies: Probabilities of monophyly, paraphyly, and polyphyly in a coalescence Zamora JC, Calonge FD, Martín MP. 2013. New sources of taxonomic infor- model. Evolution 57: 1465‒1477. mation for earthstars (Geastrum, Geastraceae, Basidiomycota): phenoloxi- Saag L, Mark K, Saag A, et al. 2014. Species delimitation in the lichenized dases and rhizomorph crystals. Phytotaxa 132: 1‒20. fungal genus Vulpicida (Parmeliaceae, Ascomycota) using gene concate Zamora JC, Calonge FD, Martín MP. 2015. Integrative taxonomy reveals an nation and coalescent-based species tree approaches. American Journal unexpected diversity in Geastrum section Geastrum (Geastrales, Basidi- of Botany 101: 2169‒2182. omycota). Persoonia 34: 130‒165. Schoch CL, Seifert KA, Huhndorf S, et al. 2012. Nuclear ribosomal internal Zardoya R, Vollmer DM, Craddock C, et al. 1996. Evolutionary conservation transcribed spacer (ITS) region as a universal DNA barcode for fungi. of microsatellite flanking regions and their use in resolving the phylogeny Proceedings of the National Academy of Sciences of the United States of of cichlid fishes (Pisces: Perciformes). Proceedings of the Royal Society, America 109: 6241‒6246. Biological Sciences 263: 1589‒1598. Schoustra SE, Debets AJM, Slakhorst M, et al. 2007. Mitotic recombination Zhang J, Kapli P, Pavlidis P, et al. 2013. A general species delimitation accelerates adaptation in the fungus Aspergillus nidulans. PLoS Genetics method with applications to phylogenetic placements. Bioinformatics 29: 3: e68 APR. 2869‒2876. Seymour FA, Crittenden PD, Wirtz N, et al. 2007. Phylogenetic and mor- Zoller S, Lutzoni F, Scheidegger C. 1999. Genetic variation within and among phological analysis of Antarctic lichen-forming Usnea species in the group populations of the threatened lichen Lobaria pulmonaria in Switzerland Neuropogon. Antarctic Science 19: 71‒82. and implications for its conservation. Molecular Ecology 8: 2049–2059. C.G. Boluda et al.: Mismatch between phenotypes and genotypes in lichenized fungi 95

Appendix 1 Chemical, geographical and morphological characters used of Bryoria sect. Implexae samples in the phenogram reconstruction (Fig. 1).

Samples Alectorialic acid Barbatolic acid Fumarprotocetraric acid Gyrophoric acid Norstictic acid Psoromic acid (1) (0) / New World Old World Thallus brownish (0) / whitish (1) Branching angles acute Branching angles obtuse Soralia ﬁssural Soralia tuberculate Pseudocyphellae inconspicuous (0) / conspicuous (1) capillaris_L01-17 1 1 0 0 0 0 0 0 1 0 0 0 0 capillaris_L06-10 1 1 0 0 0 0 0 0 1 0 0 0 1 capillaris_L07-15 1 1 0 0 0 0 0 0 1 1 0 0 0 capillaris_L08-12 1 1 0 0 0 0 0 0 1 0 0 1 0 capillaris_L13-03 1 1 0 0 0 0 0 0 1 1 0 1 0 capillaris_L14-02 1 1 1 0 0 1 0 0 1 1 0 0 1 capillaris_L141 1 1 0 0 0 0 0 0 1 0 0 1 0 capillaris_L15-15 1 1 0 0 0 0 0 0 1 0 0 0 0 capillaris_L16-21 1 1 0 0 1 1 0 0 1 1 0 0 0 capillaris_L211 1 1 0 0 0 0 0 0 1 0 0 1 0 capillaris_L270 0 1 0 0 0 0 0 0 1 0 0 1 0 capillaris_S192 1 1 0 0 0 0 0 0 1 0 0 1 0 capillaris_S2 1 1 0 0 0 0 0 0 1 0 0 1 0 friabilis_02 0 0 0 1 0 0 1 1 1 0 0 0 1 friabilis_L355 0 0 0 1 0 0 1 1 1 0 0 0 1 friabilis_L407 0 0 0 1 0 0 1 1 1 0 0 0 1 friabilis_S395 0 0 0 1 0 0 1 1 1 0 0 0 1 fuscescens_L12-03 0 0 1 0 0 0 0 1 1 0 1 1 0 fuscescens_L12-05 0 0 1 0 0 0 0 1 0 1 1 1 0 fuscescens_L139 0 0 1 0 0 0 0 1 1 0 1 1 0 fuscescens_L149 0 0 1 0 0 0 0 1 1 0 1 1 0 fuscescens_L15-21 0 0 1 0 0 0 0 1 1 1 1 1 0 fuscescens_L160 0 0 1 0 0 0 0 1 1 0 1 1 0 fuscescens_L189 0 0 1 0 0 0 0 1 1 0 1 1 0 fuscescens_L224 0 0 1 0 0 0 0 1 1 0 1 1 0 fuscescens_L232 0 0 0 0 0 0 0 1 1 0 1 1 0 fuscescens_L305 0 0 1 0 0 0 0 1 1 0 1 1 0 fuscescens_S109 0 0 1 0 0 0 0 1 1 0 1 1 0 fuscescens_S157 0 0 1 0 0 0 0 1 1 0 1 1 0 fuscescens_S24 0 0 1 0 0 0 0 1 1 0 1 1 0 fuscescens_S256 0 0 1 0 0 0 1 1 1 0 1 1 0 fuscescens_S259 0 0 1 0 0 0 1 1 1 0 1 1 0 fuscescens_S260a 0 0 1 0 0 0 1 1 1 0 1 1 0 fuscescens_S261 0 0 1 0 0 0 1 1 1 0 1 1 0 fuscescens_S267 0 0 1 0 0 0 1 1 1 0 1 1 0 fuscescens_S272 0 0 1 0 0 0 1 1 1 0 1 1 0 fuscescens_S274 0 0 0 0 0 0 1 1 1 0 1 1 0 fuscescens_S369 0 0 1 0 0 0 1 1 1 0 1 1 0 fuscescens_S379 0 0 1 0 0 0 1 1 1 0 1 1 0 fuscescens_S380 0 0 1 0 0 0 1 1 1 0 1 1 0 fuscescens_S56 0 0 1 0 0 0 0 1 1 0 1 1 0 implexa_L01-01 0 0 0 0 0 1 0 1 1 1 1 1 1 implexa_L06-05 0 0 1 0 0 1 0 1 1 0 0 1 0 implexa_L10-03 0 0 0 0 0 1 0 1 0 1 0 0 1 implexa_L11-15 0 0 0 0 0 1 0 1 0 1 0 1 0 implexa_L16-15 0 0 0 0 0 1 0 1 1 0 0 0 1 implexa_S168 0 0 0 0 0 1 0 1 1 1 1 1 1 implexa_S22 0 0 0 0 0 1 0 1 1 1 1 1 1 implexa_S36 0 0 0 0 0 1 0 1 1 1 1 1 1 implexa_S39 0 0 0 0 0 1 0 1 1 1 1 1 1 implexa_S67 0 0 0 0 0 1 0 1 1 1 1 1 1 inactiva_L206 0 0 0 0 0 0 1 0 1 0 0 0 1 inactiva_L323b 0 0 0 0 0 0 1 0 1 0 0 0 1 inactiva_L347 0 0 0 0 0 0 1 0 1 0 0 0 1 inactiva_L358 0 0 0 0 0 0 1 0 1 0 0 0 1 inactiva_S239a 0 0 0 0 0 0 1 0 1 0 0 0 1 inactiva_S384 0 0 0 0 0 0 1 0 1 0 0 0 1 inactiva_S392 0 0 0 0 0 0 1 0 1 0 0 0 1 kockiana_L394 0 0 0 0 0 1 1 0 1 1 0 0 0 kockiana_L396 0 0 0 0 0 1 1 0 1 1 0 0 0 kuemmerleana_L04-03 0 0 0 0 1 0 0 1 1 0 0 0 0 kuemmerleana_L09-04 0 0 0 0 1 0 0 1 1 1 0 0 1 kuemmerleana_L09-07 0 0 0 0 1 0 0 1 1 0 0 0 0 kuemmerleana_L16-17 0 0 0 0 1 1 0 1 1 1 1 1 1 kuemmerleana_L244 0 0 0 0 1 0 0 1 1 1 1 1 1 kuemmerleana_L274 0 0 0 0 1 0 0 1 1 1 1 1 1 kuemmerleana_L275 0 0 0 0 1 0 0 1 1 1 1 1 1 96 Persoonia – Volume 42, 2019

Appendix 1 (cont.)

Samples Alectorialic acid Barbatolic acid Fumarprotocetraric acid Gyrophoric acid Norstictic acid Psoromic acid (1) (0) / New World Old World Thallus brownish (0) / whitish (1) Branching angles acute Branching angles obtuse Soralia ﬁssural Soralia tuberculate Pseudocyphellae inconspicuous (0) / conspicuous (1) kuemmerleana_S128 0 0 0 0 1 0 0 1 1 1 1 1 1 kuemmerleana_S160 0 0 0 0 1 0 0 1 1 1 1 1 1 pikei_02 1 1 0 0 0 0 1 1 1 0 0 0 1 pikei_04 1 1 0 0 0 0 1 1 1 0 0 0 1 pikei_05 1 1 0 0 0 0 1 1 1 0 0 0 1 pikei_07 1 1 0 0 0 0 1 1 1 0 0 0 1 pikei_09 1 1 0 0 0 0 1 1 1 0 0 0 1 pikei_10 1 1 0 0 0 0 1 1 1 0 0 0 1 pikei_11 1 1 0 0 0 0 1 1 1 0 0 0 1 pikei_12 1 1 0 0 0 0 1 1 1 0 0 0 1 pikei_13 1 1 0 0 0 0 1 1 1 0 0 0 1 pikei_14 1 1 0 0 0 0 1 1 1 0 0 0 1 pikei_15 1 1 0 0 0 0 1 1 1 0 0 0 1 pikei_a 1 1 0 0 0 0 1 1 1 0 0 0 1 pikei_b 1 1 0 0 0 0 1 1 1 0 0 0 1 pikei_c 1 0 0 0 0 0 1 1 1 0 0 0 1 pikei_d 1 1 0 0 0 0 1 1 1 0 0 0 1 pikei_L197 1 1 0 0 0 0 1 1 1 0 0 0 1 pikei_L210 1 1 0 0 0 0 1 1 1 0 0 0 1 pikei_L241 1 0 0 0 0 0 1 1 1 0 0 0 1 pikei_L374 1 0 0 1 0 0 1 1 1 0 0 0 1 pikei_L376 1 0 0 1 0 0 1 1 1 0 0 0 1 pikei_L377 1 0 0 1 0 0 1 1 1 0 0 0 1 pikei_S221 1 1 0 0 0 0 1 1 1 0 0 0 1 pikei_S362 1 1 0 0 0 0 1 1 1 0 0 0 1 pikei_S368 1 1 0 0 0 0 1 1 1 0 0 0 1 pikei_S382 1 1 0 0 0 0 1 1 1 0 0 0 1 pikei_S383a 1 1 0 0 0 0 1 1 1 0 0 0 1 pikei_S390 1 1 0 0 0 0 1 1 1 0 0 0 1 pikei_S394 1 1 0 0 0 0 1 1 1 0 0 0 1 pseudofuscescens_S222 0 0 0 0 1 0 1 0 1 1 0 0 1 pseudofuscescens_S232 0 0 0 0 1 0 1 0 1 1 0 0 1 pseudofuscescens_S370 0 0 0 0 1 0 1 0 1 1 0 0 1 pseudofuscescens_S371 0 0 0 0 1 0 1 0 1 1 0 0 1 pseudofuscescens_S377 0 0 0 0 1 0 1 0 1 1 0 0 1 pseudofuscescens_S386 0 0 0 0 1 0 1 0 1 1 0 0 1 pseudofuscescens_S387 0 0 0 0 1 0 1 0 1 1 0 0 1 sp_L395 0 0 0 0 0 0 1 0 1 1 0 0 0 sp_S392 0 0 0 0 0 0 1 0 1 1 0 0 0 vrangiana_L02-20 0 0 1 0 0 0 0 1 0 1 1 1 0 vrangiana_L03-07 0 0 1 0 0 0 0 1 0 1 0 1 0 vrangiana_L05-17 0 0 1 0 0 0 0 1 0 1 0 1 0 vrangiana_L07-03 0 0 1 0 0 0 0 1 1 1 1 1 1 vrangiana_L07-19 0 0 0 0 0 0 0 1 0 1 0 0 0 vrangiana_L08-19 0 0 0 0 0 0 0 1 0 1 1 1 0 vrangiana_L08-20 0 0 1 0 0 0 0 1 0 1 0 1 0 vrangiana_L10-13 0 0 0 0 0 0 0 1 0 1 0 0 0 vrangiana_L12-11 0 0 1 1 0 0 0 1 0 1 1 1 0 vrangiana_L13-12 0 0 0 1 0 0 0 1 1 1 0 0 1 vrangiana_L272 0 0 0 1 0 0 0 1 0 1 1 1 0 vrangiana_L273 0 0 0 0 0 0 0 1 0 1 1 1 0 vrangiana_L300 0 0 0 0 0 0 0 1 0 1 1 1 0 vrangiana_L307 0 0 0 1 0 0 0 1 0 1 1 1 0 vrangiana_S10 0 0 0 1 0 0 0 1 0 1 1 1 0 vrangiana_S164 0 0 0 0 0 0 0 1 0 1 1 1 0 vrangiana_S166 0 0 1 0 0 0 0 1 0 1 1 1 0 vrangiana_S196a 0 0 0 0 0 0 0 1 0 1 1 1 0 vrangiana_S341 0 0 0 0 0 0 0 1 0 1 1 1 0 vrangiana_S385 0 0 1 0 0 0 0 1 0 1 1 1 0 vrangiana_S42 0 0 0 1 0 0 0 1 0 1 1 1 0 vrangiana_S45 0 0 0 0 0 0 0 1 0 1 1 1 0 vrangiana_S57 0 0 1 0 0 0 0 1 0 1 1 1 0 vrangiana_S59 0 0 1 0 0 0 0 1 0 1 1 1 0 vrangiana_S6 0 0 1 0 0 0 0 1 0 1 1 1 0 vrangiana_S62 0 0 0 1 0 0 0 1 0 1 1 1 0 vrangiana_S72 0 0 0 0 0 0 0 1 0 1 1 1 0 C.G. Boluda et al.: Mismatch between phenotypes and genotypes in lichenized fungi 97

Appendix 2 Phenograms based on a presence/absence distance matrix in Bryoria sect. Implexae from: a. Extrolite composition alone; b. extrolite composition, with geographical, and morphological data. ― Bold branches represent supported clades (bootstrap ≥ 70 %, approximately unbiased ≥ 95 %). ― Ale. = Alec- torialic acid; Bar. = Barbatolic acid; Fum. = Fumarprotocetraric acid; Gyr. = Gyrophoric acid; No subs. = No substances detected; Nor. = Norstictic acid; Pso. = Psoromic acid; * = Except specimen named Bryoria capillaris L14.02. 98 Persoonia – Volume 42, 2019

Appendix 3 Microsatellite fragment lengths of Bryoria sect. Implexae analysed specimens.

Sample Bi01 Bi03 Bi04 Bi05 Bi10 Bi11 Bi12 Bi14 Bi19 capillaris_L01-17 103 279 323 128 437 316 100 365 346 capillaris_L06-10 103 279 323 128 437 316 100 365 346 capillaris_L07-15 103 279 323 128 437 316 100 365 346 capillaris_L08-12 103 279 323 128 434 316 103 361 346 capillaris_L13-03 94 279 325 138 434 318 115 361 346 capillaris_L14-02 112 279 327 128 437 320 100 365 346 capillaris_L141 123 281 323 136 434 316 100 361 346 capillaris_L15-15 112 279 323 – 437 316 100 365 346 capillaris_L16-21 103 279 323 128 437 316 100 365 346 capillaris_L211 112 279 323 136 434 316 103 361 346 capillaris_L270 112 279 323 138 434 316 124 361 346 capillaris_S192 112 279 323 128 437 316 100 365 346 capillaris_S2 94 279 323 138 434 316 124 361 346 friabilis_02 114 281 317 137 434 310 100 – 350 friabilis_L355 120 277 316 137 436 310 100 365 350 friabilis_L407 132 281 316 137 434 310 131 365 346 friabilis_S395 109 277 316 132 438 310 100 365 350 fuscescens_L12-03 112 279 323 138 434 316 103 361 352 fuscescens_L12-05 94 281 323 138 434 316 103 361 350 fuscescens_L139 94 279 323 138 434 316 103 361 352 fuscescens_L149 123 279 323 138 434 316 103 361 350 fuscescens_L15-21 123 277 325 138 434 318 115 361 352 fuscescens_L160 94 281 323 136 434 316 103 361 352 fuscescens_L189 112 279 323 138 434 316 100 361 352 fuscescens_L224 112 279 323 138 434 316 103 361 352 fuscescens_L232 94 279 323 138 434 316 103 361 352 fuscescens_L305 117 279 323 138 434 316 137 361 352 fuscescens_S109 112 281 323 138 434 316 118 361 350 fuscescens_S157 117 279 323 136 434 316 103 361 350 fuscescens_S24 112 281 323 138 434 316 106 361 352 fuscescens_S256 94 281 323 – 434 316 124 363 354 fuscescens_S259 94 277 323 138 437 316 103 363 352 fuscescens_S260a 82 279 323 138 437 316 103 363 352 fuscescens_S261 82 279 323 138 437 316 103 363 352 fuscescens_S267 82 279 323 138 437 316 103 363 352 fuscescens_S272 82 279 323 138 437 316 103 363 352 fuscescens_S274 94 279 323 138 434 316 – 361 350 fuscescens_S369 – 279 323 138 437 316 103 363 352 fuscescens_S379 94 279 323 138 437 316 100 363 352 fuscescens_S380 112 279 323 138 434 316 118 361 352 fuscescens_S56 123 279 323 136 434 316 100 361 350 glabra_01 114 283 306 132 426 299 115 369 344 glabra_02 120 283 306 132 426 299 – 369 344 glabra_03 120 283 306 132 426 299 – 369 344 glabra_04 120 283 306 132 426 299 – 369 344 glabra_05 120 283 306 132 426 299 – 369 344 glabra_L186 114 283 304 132 426 297 115 371 344 glabra_L406 114 283 304 132 426 297 115 371 344 glabra_L414 114 283 306 132 426 299 115 371 344 glabra_S388 114 283 306 132 426 299 115 371 344 implexa_L01-01 135 281 323 138 434 316 106 361 350 implexa_L06-05 112 263 323 136 434 316 103 361 350 implexa_L10-03 106 279 323 128 437 316 100 365 346 implexa_L11-15 94 279 323 136 435 316 103 361 352 implexa_L16-15 112 281 325 138 434 318 115 361 352 implexa_S168 112 279 323 138 434 316 115 361 350 implexa_S22 94 279 323 136 436 316 118 361 352 implexa_S36 94 279 325 136 434 318 103 361 346 implexa_S39 – – 323 138 434 316 – 361 350 implexa_S67 94 279 325 136 434 318 103 361 346 inactiva_L206 114 277 317 137 434 311 100 365 350 inactiva_L323b 114 281 316 132 434 310 131 365 350 inactiva_L347 114 277 317 132 434 310 124 365 350 inactiva_L358 109 281 317 132 436 310 106 365 350 inactiva_S239a 109 277 314 132 434 308 100 365 350 inactiva_S384 114 277 316 137 434 310 100 365 350 inactiva_S392 120 281 316 137 434 310 100 365 350 kockiana_L394 94 279 317 136 472 310 109 365 344 kockiana_L396 94 279 317 136 472 310 109 365 344 kuemmerleana_L04-03 129 279 325 138 434 318 103 361 352 kuemmerleana_L09-04 112 281 323 128 437 316 100 365 346 kuemmerleana_L09-07 112 281 323 128 437 316 100 365 346 kuemmerleana_L16-17 112 281 323 138 434 316 106 361 352 kuemmerleana_L244 117 279 323 138 434 316 115 361 350 kuemmerleana_L274 94 279 323 136 434 316 103 361 352 kuemmerleana_L275 94 279 323 136 434 316 103 361 352 kuemmerleana_S128 100 279 323 136 434 316 100 361 354 kuemmerleana_S160 117 279 323 138 434 316 103 361 350 C.G. Boluda et al.: Mismatch between phenotypes and genotypes in lichenized fungi 99

Appendix 3 (cont.)

Sample Bi01 Bi03 Bi04 Bi05 Bi10 Bi11 Bi12 Bi14 Bi19 pikei_02 117 277 317 137 434 310 100 365 344 pikei_04 117 277 317 137 434 310 100 365 344 pikei_05 117 277 317 137 434 310 100 365 344 pikei_07 117 281 317 137 434 311 109 – 344 pikei_09 114 281 317 132 – 310 100 – 344 pikei_10 117 277 317 137 434 311 100 365 344 pikei_11 114 277 317 137 434 310 100 365 344 pikei_12 114 277 – 137 434 – 100 – 344 pikei_13 – 277 317 137 434 310 118 365 344 pikei_14 117 277 317 137 434 310 127 365 344 pikei_15 114 281 317 137 436 310 109 – 344 pikei_a 114 – 323 132 437 316 100 – 346 pikei_b 117 277 317 137 434 311 100 365 344 pikei_c 108 – – 137 – 314 100 326 344 pikei_d 117 277 – 137 437 310 100 – 344 pikei_L197 114 277 316 132 434 310 109 365 344 pikei_L210 109 277 316 132 434 310 100 365 344 pikei_L241 126 277 316 137 434 310 109 – 352 pikei_L374 114 277 316 137 434 310 106 365 344 pikei_L376 132 277 316 137 434 310 127 365 352 pikei_L377 109 277 316 137 434 310 103 365 344 pikei_S221 109 277 316 132 436 310 106 365 352 pikei_S362 114 281 317 137 434 311 100 365 344 pikei_S368 112 281 316 132 436 310 109 365 344 pikei_S382 114 281 317 137 434 311 100 365 344 pikei_S383a 114 281 317 132 436 311 100 365 344 pikei_S390 114 277 316 137 434 310 100 365 344 pikei_S394 126 277 316 137 434 310 100 365 352 pseudofuscescens_S222 – 277 316 132 – 310 100 365 – pseudofuscescens_S232 114 277 317 137 436 310 100 365 350 pseudofuscescens_S370 117 277 316 132 434 310 100 365 350 pseudofuscescens_S371 117 281 316 132 434 310 100 365 350 pseudofuscescens_S377 – 277 – 132 434 311 100 365 – pseudofuscescens_S386 112 277 317 132 438 311 100 365 350 pseudofuscescens_S387 117 277 316 132 434 310 127 365 350 sp_L395 94 273 317 136 460 310 109 365 344 sp_S392 97 283 317 136 472 310 118 365 344 vrangiana_L02-20 112 279 323 138 434 316 106 361 352 vrangiana_L03-07 94 279 325 136 434 318 115 365 346 vrangiana_L05-17 117 281 323 138 434 316 121 361 350 vrangiana_L07-03 94 281 323 138 434 316 118 361 352 vrangiana_L07-19 123 279 323 138 434 316 103 361 350 vrangiana_L08-19 94 281 323 136 436 316 103 361 352 vrangiana_L08-20 94 279 323 138 434 316 121 361 352 vrangiana_L10-13 123 283 323 128 437 316 100 365 346 vrangiana_L12-11 94 279 323 138 434 316 144 361 352 vrangiana_L13-12 94 279 323 138 434 316 103 361 352 vrangiana_L272 112 279 323 136 434 316 – 361 350 vrangiana_L273 94 279 323 138 436 316 103 361 350 vrangiana_L300 94 281 323 136 434 316 115 361 350 vrangiana_L307 123 279 325 138 434 318 103 361 352 vrangiana_S10 123 279 323 136 434 316 124 361 352 vrangiana_S164 94 279 323 136 434 316 127 361 350 vrangiana_S166 117 279 323 138 434 316 144 365 352 vrangiana_S196a 94 281 323 138 434 316 118 361 350 vrangiana_S341 – 279 323 – 434 316 – 361 350 vrangiana_S385 94 279 323 138 434 316 103 361 350 vrangiana_S42 117 279 323 138 434 316 131 361 352 vrangiana_S45 112 279 323 136 434 316 100 361 350 vrangiana_S57 94 279 323 138 434 316 115 361 352 vrangiana_S59 94 279 323 138 434 316 115 361 352 vrangiana_S6 123 279 323 138 434 316 103 361 350 vrangiana_S62 112 279 323 138 434 316 115 361 352 vrangiana_S72 94 279 323 136 436 316 103 361 352 100 Persoonia – Volume 42, 2019

Appendix 4 Haplotype network in Bryoria sect. Implexae of a concatenated matrix containing ITS, IGS and GAPDH sequences. The analysis coded gaps as missing data and used a 95 % connection limit. Numbers represent the specimens shown in Table 1 and colours depict the STRUCTURE microsatellites genepool (Fig. 2). Connecting line length do not depict the genetic distance. Each line represents a single mutation connected by black small circles. Circle size is related with the number of analysed specimens. ― * = WDb (Wide Distributed brown cluster) specimens.

Appendix 5 Distribution of Bryoria sect. Implexae specimens examined. Two samples of geographical interest, not analysed in this study, have been added: Bryoria kockiana (Canada, British Columbia, 1982, Goward 82-1040, UBC – paratype; cf. Velmala et al. 2014) and Bryoria fuscescens (Tanzania, Kilimanjaro, 2016, Boluda & Kitara, MAF-Lich. 20750). ― Basemap source: U.S. National Park Service (NPS) Natural Earth physical map.